SEARCH

SEARCH BY CITATION

Keywords:

  • Emotions;
  • Modules;
  • Modularity;
  • Is-ought;
  • Intuition;
  • Moral;
  • Mencius

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction: Moral modularity
  4. 2. Ancient modularity
  5. 3. Twenty-first century modularity
  6. 4. Classical and contemporary modularity
  7. 5. Modules with and without emotional oomph
  8. 6. Darwin and moral modularity
  9. 7. Normativity
  10. 8. Normativity and modularity
  11. References

Flanagan (1991) was the first contemporary philosopher to suggest that a modularity of morals hypothesis (MMH) was worth consideration by cognitive science. There is now a serious empirically informed proposal that moral competence is best explained in terms of moral modules-evolutionarily ancient, fast-acting, automatic reactions to particular sociomoral experiences (Haidt & Joseph, 2007). MMH fleshes out an idea nascent in Aristotle, Mencius, and Darwin. We discuss the evidence for MMH, specifically an ancient version, “Mencian Moral Modularity,” which claims four innate modules, and “Social Intuitionist Modularity,” which claims five innate modules. We compare these two moral modularity models, discuss whether the postulated modules are best conceived as perceptual/Fodorian or emotional/Darwinian, and consider whether assuming MMH true has any normative ethical consequences whatsoever. The discussion of MMH reconnects cognitive science with normative ethics in a way that involves the reassertion of the “is-ought” problem. We explain in a new way what this problem is and why it would not yield. The reason does not involve the logic of “ought,” but rather the plasticity of human nature and the realistic options to “grow” and “do” human nature in multifarious legitimate ways.


1. Introduction: Moral modularity

  1. Top of page
  2. Abstract
  3. 1. Introduction: Moral modularity
  4. 2. Ancient modularity
  5. 3. Twenty-first century modularity
  6. 4. Classical and contemporary modularity
  7. 5. Modules with and without emotional oomph
  8. 6. Darwin and moral modularity
  9. 7. Normativity
  10. 8. Normativity and modularity
  11. References

In Varieties of Moral Personality, Owen Flanagan (1991) argued that an hypothesis of the modularity of morals, MMH, was worth serious consideration by cognitive science on the grounds that (some aspects of) morality seems to be adaptive, possibly even a biological adaptation, and in addition morality seems to involve multifarious competencies suited to different social ecologies (or different aspects of a single social ecology) rather than a unitary competence. Specifically, the argument was that virtues such as justice and benevolence have different emotional bases, domains, and learning histories, and thus possess characteristics of other skills that have been profitably modeled modularly, such as face recognition, language, the senses, and the basic emotions.

There is now, a decade and a half later, a serious empirically informed proposal put forward by social intuitionists, which claims that moral competence and moral performance are best explained in terms of intuitional moral modules.1 Social intuitionism is the name for the view that the structure of moral character and its components—moral habits, virtues, and the like—are best explained as a cultured elaboration of a set of semiautonomous “foundations,” which were designed in the first instance (by natural selection) to meet common adaptive challenges faced by Homo sapiens. The sprouts or “foundations” in our nature that are culturally elaborated to form “character,” which John Dewey called “the interpenetration of habits,” name evolutionarily ancient, fast-acting, psychological (perception-affect-action) programs that constitute what we might call “first nature.”2MMH as put forward by the social intuitionists fleshes out an idea that was anticipated, on various interpretations, in Aristotle, Mencius, and Darwin, and which in its current form is supported by interdisciplinary work in anthropology, cross-cultural psychology, primatology, and economics (Brown, 1991; Cosmides & Tooby, 1992; Fiske, 1991, 1992, 2004; Schwartz & Bilsky, 1990; Shweder & Haidt, 1993; Shweder, Much, Mahapatra, & Park, 1997; de Waal, 1991, 1996).

Methodologically, MMH utilizes an approach that Howard Gardner (1983, 1993, 2006) deployed in his theory of “multiple intelligences” to open up the concept of intelligence to a more anthropologically realistic and ecologically valid account of what “intelligence” means (possibly, if one is a realist about psychological kinds, what intelligence is) than simply what IQ tests test. The idea is to study cross-culturally the aspects of mind and life that are deemed to involve “intelligence” (or in the current case “morals”) broadly construed, rather than accepting a culturally and normatively specific conception that privileges the aspects or kinds of intelligence (or morality) that some tradition endorses or favors, and thus which it, not surprisingly but incorrectly, claims captures what “intelligence” (or “morality”) really is. In Gardner’s (1983, 8) case, the evidence he suggests provides “persuasive evidence of several relatively autonomous human intellectual competences.” Moral modularity promises something similar: Moral competence consists of, or is the emergent product of, a set of autonomous or relatively autonomous sociomoral competences. One appealing feature of the social intuitionists’ version of MMH is that it claims to offer a universal psychosocial baseline for comparing and contrasting moral orientations across individuals and cultures. Depending on how MMH is framed, it might also be read as embedding a criterion or criteria for judging the adequacy of a type or level of moral competence and performance.3

In this paper, we discuss two different versions of MMH: one from classical Chinese philosophy, specifically from Mencius 5th c. BCE, which we call Mencian moral modularity (MMM); the other from 21st century social psychology, which we call social intuitionist modularity (SIM). After sketching and critically comparing MMM and SIM, we ask this question: Assuming that some version of MMH is true—possibly MMM or SIM—does this have any normative ethical consequences whatsoever? Can anything about how we ought morally to perceive, feel, think, and act be extracted or derived from the claims about universal modules that are keyed to particular kinds of situations? The first question is central to cognitive science, the second to ethics.

2. Ancient modularity

  1. Top of page
  2. Abstract
  3. 1. Introduction: Moral modularity
  4. 2. Ancient modularity
  5. 3. Twenty-first century modularity
  6. 4. Classical and contemporary modularity
  7. 5. Modules with and without emotional oomph
  8. 6. Darwin and moral modularity
  9. 7. Normativity
  10. 8. Normativity and modularity
  11. References

Mencius (5th c. BCE), the most famous classical Chinese philosopher after Confucius, differed from the Master in this way: Confucius described the virtuous person as ren—a humane person—where ren ascribes virtue generally. For Mencius, on the other hand, ren is a specific virtue, benevolence, one of a team of (at least) four, which together constitute virtue or good character. Mencius claims that virtue comes from enhancing or growing four innate sprouts:

Humans all have hearts that are not unfeeling toward others. Suppose someone suddenly saw a child about to fall into a well: everyone in such a situation would have a feeling of alarm and compassion—not because one sought to get in good with the child’s parents, not because one wanted fame among their neighbors and friends, and not because one would dislike the sounds of the child’s cries. [F]rom this we can see that if one is without the heart of compassion, one is not a human. If one is without the heart of deference, one is not a human. The heart of compassion is the sprout of benevolence. The heart of disdain (shame/disgust) is the sprout of righteousness. The heart of deference is the sprout of propriety. The heart of approval and disapproval is the sprout of wisdom. (Mencius, 2008, 2A6, see also 6A6)

Next, Mencius says this: “People having these four sprouts is like their having four limbs.” Later at 4A27, he writes that if one grows all four moral sprouts, all four limb buds, “then without realizing it one’s feet begin to step in time to them and one’s hands dance according to their rhythms.”

From these passages we can extract what is arguably the first text known, East or West, to express a version of MMH, actually two versions, a descriptive and a normative version. Call these Mencian moral modularity (MMM).

MMMDescriptive Human nature contains sprouts for four different moral competencies.

MMMNormative Moral excellence involves growing all four sprouts to maturity.

The descriptive thesis—MMMDescriptive—tells us that human nature contains the sprouts of compassion, shame/disgust, deference, and distinguishing right from wrong each of which, to speak in an Aristotelian way, has a trajectory, a directionality, a proper function, an ergon, a potential that it seeks to actualize. These four spouts mature into the four cardinal virtues of benevolence (ren), righteousness (yi), propriety (li), and wisdom (zhi).

The normative thesis—MMMNormative—says that growing all four is good, something we ought to do. Not doing so would be like being a person with missing or lost limbs. Assuming one grows the sprouts properly, one is truly human and a good or decent person. One can fail to be fully human if any one of the sprouts lies latent or dies (Van Norden, 2007). Just as the loss or failure to grow any of the four limbs would lead to difficulty moving through space, loss of or failure to grow any of the four Mencian sprouts will lead to difficulty negotiating sociomoral space; one will not acquire the ability to “dance” in the effortless and graceful (wu-wei) way that a morally well-formed person does.4

Mencius’s modularity thesis, MMM, can be summarized as follows: Human nature contains four sprouts or hearts that in a normal environment grow into four distinct virtues that taken together constitute good character or virtue, generally. Sympathy (of the sort that is activated by children falling into wells) is the sprout for benevolence (ren); shame and disgust for righteousness (yi); deference for propriety (li); a sense of true and false, accurate and inaccurate, match and mismatch, approval and disapproval is the sprout for practical wisdom (zhi), of appraising persons and situations for who and what they really are, for knowing what to do, and when and how to do it, and so on. Each sprout is an innate cognitive-affective-conative disposition that possesses the potential, the natural trajectory to grow into one of the four cardinal virtues. If these sprouts are planted in a normal environment, they will grow like the four limbs do. If they receive suboptimal nourishment, they will grow some; and if they are not nourished at all, they will not grow (6A8). The best outcome is that the four sprouts blossom into the four cardinal virtues. Barring congenital abnormality or abnormality in social conditions the best outcome is realized and virtuous agents emerge.5

So Mencius defends both MMMDescriptive and MMMNormative. Although no one ever asks why it is a good thing that we have four limbs and why they are sized the way they are, it is instructive to consider what one could say if asked. One way to defend Mencian normative modularity for limbs would be by claiming that:

  • 1
    Evolution settled on a four leg/four limb design because it was an adaptation = adaptationhistorical
  • 2
    This design is still adaptive = adaptationcurrent ecology6
  • 3
    This four limb design emerges naturally in a species universal way across normal ecologies; and thus,
  • 4
    We ought to grow our arms and legs the way nature designed them to grow.

The “ought” in (4) expresses the bidirectional agent-to-world goodness-of-fit between a universal phenotypic trait and the world (literally, the earth). If MMMNormative were credible, the parallel would run as follows:

  • 1
    Evolution settled on four moral modules (=sprouts) because they were adaptations = adaptationhistorical
  • 2
    They are still adaptive = adaptationcurrent ecology
  • 3
    The modules (=sprouts) emerge, grow, and are tuned (roughly) the same way across all natural and social ecologies; and thus,
  • 4
    We ought to grow the modules the way Mother Nature (Tian = heaven) designed them to grow.

3. Twenty-first century modularity

  1. Top of page
  2. Abstract
  3. 1. Introduction: Moral modularity
  4. 2. Ancient modularity
  5. 3. Twenty-first century modularity
  6. 4. Classical and contemporary modularity
  7. 5. Modules with and without emotional oomph
  8. 6. Darwin and moral modularity
  9. 7. Normativity
  10. 8. Normativity and modularity
  11. References

Jonathan Haidt and his colleagues (Haidt, 2001; Haidt & Graham, 2007; Haidt & Joseph, 2004, 2007) have proposed a version of MMH that is advertised as a social intuitionist model to convey that the modules consist of dispositions to have rapid-fire cognitive-affective-conative reactions (the intuitions) that are triggered by specific types of social or environmental situations, akin to Mencius’s example of the universal human impulse to save the child falling into the well (see Table 1). Haidt (2001) argues that while people commonly feel, or say, that they reason in making moral judgments, it is sometimes, even often, the case that one’s firm and confident moral judgments co-occur with limbic system activation, and thus in all likelihood supervene on it, and that therefore the “reasoning” which supervenes on PFC activity is post hoc rationalization of prior intuitions. We refer to the model promoted by Haidt and his colleagues as the social intuitionist modularity model or SIM, for short. SIM is one version of MMH.

Table 1.    The five foundations of intuitive ethics
 Harm/CareFairness/ReciprocityIngroup/LoyaltyAuthority/RespectPurity/Sanctity
  1. Source. From Haidt and Joseph (2007, p.392). We thank Oxford University Press and the authors for permission to reprint the table here.

Adaptive challengeProtect and care for young, vulnerable, or injured kinReap benefits of dyadic cooperation with non-kinReap benefits of group cooperationNegotiate hierarchy, defer selectivelyAvoid microbes and parasites
Proper domain (adaptive triggers)Suffering, distress, or threat to one’s kinCheating, cooperation, deceptionThreat or challenge to groupSigns of dominance and submissionWaste products, diseased people
Actual domain (the set of all triggers)Baby seals, cartoon charactersMarital fidelity, broken vending machinesSports teams one roots forBosses, respected professionalsTaboo ideas (communism, racism)
Characteristic emotionsCompassionAnger, gratitude, guiltGroup pride, belongingness; rage at traitorsRespect, fearDisgust
Relevant virtues [and vices]Caring, kindness [cruelty]Fairness, justice, honesty, trustworthiness vv[dishonesty]Loyalty, patriotism, self-sacrifice [treason, cowardice]Obedience, deference [disobedience, uppitiness]Temperance, chastity, piety, cleanliness [lust, intemperance]

Social intuitionist modularity draws on interdisciplinary work (Brown, 1991; Fiske, 1991, 1992, 2004; Schwartz & Bilsky, 1990; Shweder & Haidt, 1993; Shweder et al., 1997; de Waal, 1991, 1996) and claims that there are five types of social situations that people everywhere evaluate in affectively loaded moral terms:7

  • 1
    Harm/care
  • 2
    Fairness/reciprocity
  • 3
    Authority/respect
  • 4
    Purity/sanctity
  • 5
    Ingroup/loyalty

Evaluative intuitions in these domains are found cross-culturally among humans and possibly in nonhuman primates as well (e.g., Brosnan & de Waal, 2003 claim anger to unfair rewards is found in capuchins). Social intuitionist modularity can be then defined this way:

SIMDescriptiveHomo sapiens possess (at least) five innate intuitional psychological modules that are activated in normal social environments, can be grown, and are the basis of morality.

The key ideas that define SIM are that there are (at least) these five intuitional modules and that something in the vicinity of virtues, or special purpose moral skills, are built upon them. These five dispositional mechanisms underwrite complex multidirectional syndromes (mind-world-action) that arose to meet specific adaptive challenges and that serve as the foundation of morality, or something in the vicinity. The outcome of building morals on modules might be that we understand moral agency in terms of the application of distinctive domain-specific moral skills or as what Dewey called the “interpenetration of habits.”

In philosophy, virtues, as special purpose moral skills, are defined as dispositions to perceive, feel, judge, and act in a way that is responsive to tokens of a situation type (Flanagan, 1991, 2009). Courage, temperance, benevolence, and so on are virtues that are appropriately activated by situations that call for them. A kind person sees the old lady standing on the subway and gives her his seat. A decent person feels sympathy for the child who scrapes her knee and goes to help her. A courageous person sees when the rights of the powerless are being trampled and stands up for them (even at cost to herself). In the normal life of a virtuous person, declarative rules are not normally consulted and need not be consulted in cases such as these.8 The virtuous person unlike what Aristotle called the “continent person” (whom Kant admired) moves in that wu-wei (effortless) manner that Mencius celebrates as suited to our kind of animal.

Social intuitionist modularity mirrors Mencius’s model in making both a descriptive (there are five innate sprouts) and a normative claim (minimally a claim that we ought to recognize the importance of all five to morality and that doing so will improve moral comprehension across life forms; possibly something stronger, to the effect that normal moral competence involves growing all the modules and coordinating them).9 Defenders of SIM, however, are sensibly cautious about offering anything as strong as the Mencian limb analogy. They point out that which sprouts develop, and the extent to which different cultures or subcultures build on the modules, depend on environmental and social inputs. Privileging some sprouts at the expense of others does not necessarily prevent one from being virtuous, as “cultures vary to the degree to which they build virtues on these five foundations” (Haidt & Graham, 2007, p. 99). Nonetheless, “the available range of human virtues is constrained by the five sets of intuitions that human minds are prepared to have” (Haidt & Graham, 2007, p. 106). This last point, if true, is important. It would mean that “morality” as a psychological kind is restricted to the original modules and extensions of the modules.10 The latter is compatible with metaphor and metonymy resulting in what is called “moral” or “morality” having wider scope than human psychology “pulls for,” which could possibly explain what is going on when people or communities dispute questions about whether some problem is or is not a moral problem.

In any case, SIM offers a modified nativist theory. The modules as originally set (or maturationally programmed to emerge in a normal environment) constitute the initial settings both in terms of which basic emotions are activated by which situations and how high or low the emotional responses are tuned. Extensions of the range of activation of the modules and the tuning up or down of the strength of emotional response, and actions taken, are accounted for by culture. Moral differences at every level, between individuals, across cultures, subcultures, and so on, are explained by differences in the degree to which the five modules are tuned, what situations they are tuned to, and the relative priority given to the various modules. Thus, from this shared set of five intuitions (modules), various cultures develop moralities that extend (or suppress) the original modules in different ways.

Because SIM emphasizes the innate emotional bases of moral response, virtue theory is taken to be more consistent with the empirical findings than are other philosophical models of moral psychology. Virtues are dispositions to respond perceptually, affectively, and in action in quick, domain specific ways, whereas a rule-based theory such as utilitarianism or Kantianism is couched in terms of consultation with and application of a general-purpose rule, the principle of utility or the categorical imperative. According to SIM, although the five foundational modules11 underwrite virtues, they are not themselves virtues but are “essential tools in the construction of virtues” (Haidt & Joseph, 2007, p. 63). Thus, we can think of the modules as potential virtues. One interesting question is whether the modules also provide the sprouts upon which vice grows or can grow (think of the desire to harm another who harms me as the basis for Schadenfreude). One could of course be skeptical that there is any real moral force to designations of virtue and vice. All naturalistic accounts of morality can be taken (although it is not necessary) to show that such designations as “virtue” and “vice” are to be read as honorifics (or pejoratives) pinned on different ways of cultivating the modules. Different societies favor different extensions of the modules. There is no deep answer to the question of which way is the right way.

The question remains whether developing all five foundations to some small or large degree is typical or, what is different, necessary for an adequate morality. If SIM is true, then the surface structural differences among moralities, about which most moral disagreements turn, are different expressions of the same underlying deep structure.12 Furthermore, the evidence suggests that either some cultures build morality on a subset of these five foundations or, more likely, that if and insofar as all five intuitional modules are appealed to in all societies, they are either hooked up with different domains of activation, are tuned up-down (higher and lower) in response to different activating conditions, or, commonly, both.

4. Classical and contemporary modularity

  1. Top of page
  2. Abstract
  3. 1. Introduction: Moral modularity
  4. 2. Ancient modularity
  5. 3. Twenty-first century modularity
  6. 4. Classical and contemporary modularity
  7. 5. Modules with and without emotional oomph
  8. 6. Darwin and moral modularity
  9. 7. Normativity
  10. 8. Normativity and modularity
  11. References

There are two issues that need to be addressed: First, is SIM Mencian in substance, in content, or in function, where by “function” we mean “proper evolutionary function” (ergon for Aristotle)? Second, are the SIM modules Mencian insofar as it is desirable, that is, normative, to grow and develop all of them, as it is, for example, to grow and develop all four limbs? To answer these questions it will be helpful to line up the two sets of modules, classical MMM and contemporary SIM (Table 2).

Table 2.    Relationship between Mencian modules and social intuitionist modules
MMMSIM
Sympathy-benevolence (ren)Harm/care
Deference-propriety (li)Authority/respect
Shame/disgust-righteousness (yi)Purity/sanctity
Approval/disapproval-wisdom (zhi)Fairness/reciprocity
 Ingroup/loyalty

Assuming that moral modules (“sprouts” in Mencius) pick out a small set of universal phenotypic traits that are adaptationshistorical one might expect the MMM list and the SIM list to be the same. The reason is that smart people can see adaptations without knowing anything about the Darwinian theory that explains what an adaptation is, how one works, and so on. But, the lists differ and not just because Mencius has only four, not five sprouts as SIM does. Mencius does not have a sprout for justice/fairness. This, of course, does not mean that Mencius does not recognize some such moral competence or universal feature of moral life (showing that he does or does not would require deep textual exegesis, which we do not engage in here). That said, justice as fairness is not marked off and treated much nor is it considered a major virtue in classical Chinese philosophy. However, if there is a justice/fairness sprout in human nature (as there may be among capuchins and canines [Brosnan & de Waal, 2003]) but it shows up neither as a sprout nor a virtue in MMM, we are owed an explanation as to why it was not seen in the 5c. BCE by Mencius.

One might look for justice/fairness in Mencian righteousness (yi), but this will be difficult since the virtue of yi is rooted in the sprout of purity/shame, which according to SIM is a whole different deal—it supports not judgments of fairness but judgments about whether marrying first cousins or being an atheist makes one dirty or yucky. Furthermore, although MMM has an up-down module in “deference,” the sprout for propriety, which maps nicely on to the authority/respect module of SIM, there is no sprout that maps onto the ingroup/loyalty module of SIM. Then, again, one could make a plausible (but not decisive) argument to the effect that the Chinese tradition makes much ado about filial piety (xiao), which is a paradigm case of an ingroup virtue that starts with one’s parents and older siblings and then generalizes to other elders inside one’s culture.13 This, if true, might lay the basis for an argument to the effect that the ingroup/loyalty module is a subspecies of authority/respect, or vice versa.14 But it will take work, and all these mapping problems might make one worry about the intuitiveness of the intuitional modules themselves.

Furthermore, when the two lists of modules, MMM and SIM, are lined up, Mencian “wisdom” (zhi) stands out as a loner and the reason is informative. Mencian wisdom (zhi) appears to be a largely cognitive meta-skill akin to Aristotelian phronesis, practical wisdom, which involves the abilities to read other people’s character, to skillfully coordinate means and ends, to apply a principle of the mean, and so on (Van Norden, 2007, p. 123). Wisdom (zhi) is not like the other modules because it does not have the property of being rooted in the emotions, being fast acting, or automatic.15 But one can see its usefulness: A standard difficulty for virtue theories is how to break ties in a system that consists of virtues all the way up and all the way down. What should I do when I am called upon to be just and compassionate at the same time or when my powerful (purity) desire to ostracize or crush the slimy scum bag conflicts with my impulse to be compassionate? Which is trump?

It is not clear how modules, even when culturally elaborated, solve such problems among themselves except by sheer strength. A practical general reasoning ability (zhi) could help here, especially if it held a trump rule to the effect that if/when there is conflict between, say, the purity and compassion modules, the latter (or former) is trump. Indeed, a close reading of Mencius indicates that he in fact endorses such trump rules. For Mencius, benevolence (ren) and righteousness (yi) are the two most important virtues, more important than propriety (li) (Van Norden, 2007, p. 273).16 How did Mencius, or does anyone, gain this ranking hierarchy? The best answer is that the hierarchy among his intuitional modules, now three (subtracting wisdom [zhi]), is discovered or seen by wisdom (zhi). And if this is right (and it is), then the whole Mencian analogy between the four sprouts and the four limbs starts to come undone or at least begins to show the limits of its usefulness. If wisdom (zhi) is a meta-skill, a principle, a heavily cognitive competence, or a set of skills for utilizing the core cardinal virtues (now three) and if benevolence and righteousness trump propriety in importance, then the analogy with growing and coordinating all four limbs breaks down; unless, that is, some limbs are to be stronger (or longer) than others and one limb (wisdom) is less like a limb than like the mind. Wisdom on the meta-skill or rule view is more like motor cortex or even PFC (prefrontal cortex), which controls the limbs, via motor areas, than like one of the limbs themselves (now three).

If one attraction of the MMM is that it seems to maintain a smooth relation between “is” and “ought,” between description and normativity, then in fact upon reflection it does not do this. If another attraction of MMM is that it anticipates the modern modularity view, SIM, then in fact in does not do that either. There is not a match between the list of philosophy’s first great moral modularist and its 21st century mate—between MMM and SIM. In the end, Mencius is not a full-fledged modularist, because he sees the need for—or even if he does not see the need for it, he imports—a meta-skill, namely wisdom (zhi), for cognitive control or orchestration of the first-order virtues, whatever the number and nature of these might be. This can be plausibly read as an important, even if inadvertent, insight about the need for and role of nonmodular elements for successfully negotiating the sociomoral domain in most actual worlds. The upshot is that MMM does not succeed, upon close scrutiny, in describing moral competence as fully modular, nor does a credible moral analog of the normative four-limb analogy emerge.

Putting aside the special problem due to the nonmodular aspects of Mencian wisdom (zhi), one way of explaining the lack of fit between the two lists, or of explaining it away, requires particular attention to the normativity issue. Perhaps Mencius was noticing the sprouts that his culture valued, idealized, and amplified but not all the sprouts available for valuing or amplifying, in which case Mencius makes no contribution to ethics as such—if there be such a thing—but only to ancient Chinese cultural anthropology or ancient Chinese moral psychology. Indeed one might worry about Mencius’s ability to see first nature in a neutral way, since his sprouts are pretty much exactly what one should expect if Mother Nature had given us the settings to be Confucians, as opposed to Buddhists. This suggests the possibility that if there are moral modules and they are comparable to limbs, then they are comparable to the limbs of a creature that has no determinate number and size of limbs but some range that constitutes normalcy. Imagine that an octopus, now a “ploctopus,” can have anywhere from four to twelve limbs depending on the local ecology. The right number is determined relationally in terms of what number of limbs is best suited to the environment or, what is utterly different, what number the environment happens to serendipitously select for. Because the “right way” to grow the sprouts is determined by a dynamic relationship between the individual or group and the environment, the modules lack the teleological directedness of the Aristotelian virtues.17

A similar problem arises when considering SIM’s modules as both descriptive and normative, as containing information about both what the innate moral dispositions are and what the extended competence is supposed to be. Recent work indicates that the morality of American Liberals rests primarily on considerations of harm and fairness, whereas the morality of American Conservatives rests on considerations corresponding to all five SIM modules (Haidt & Graham, 2007, p. 13). Cultures or, as in this case, even subcultures “can specialize in a subset of human moral potential” (Haidt, 2001, p. 827). Haidt and his colleagues are not explicit about whether the difference between Liberals and Conservatives is one of privileging certain modules over others when there are conflicts, whether some modules are simply set relatively lower or higher in one group than the other, or whether (for Liberals) some modules are tuned so low as to be functionally in the “off” position. According to SIM, divergent conceptions of the appropriate boundary conditions (domains) of the modules, divergent relative priority among the modules, or over- or underdevelopment of the modules would each deliver us divergent moralities, which are nonetheless built from the same suite of possible modules. So what is the right way to grow and/or tune the modules? If the answer is not in the modules (and it is not, because they lack the self-directedness of Aristotelian virtues), where—if anywhere—is it? We return to this question in the final section. But first, we try to explain more precisely what kind of modules moral modules are if, that is, there are any.

5. Modules with and without emotional oomph

  1. Top of page
  2. Abstract
  3. 1. Introduction: Moral modularity
  4. 2. Ancient modularity
  5. 3. Twenty-first century modularity
  6. 4. Classical and contemporary modularity
  7. 5. Modules with and without emotional oomph
  8. 6. Darwin and moral modularity
  9. 7. Normativity
  10. 8. Normativity and modularity
  11. References

All modules are Darwinian. That is, if there are any modules at all, they consist of universal phenotypic traits that have accrued to our type of animal, that have accumulated to our dasein, through the long meticulous process of natural selection and now come with the equipment. Anatomy is a modular field par excellence and the possibility proof that a field can study complex integrated systems modularly.18 But the modules anatomy depicts—kidneys, the liver, and so on—are not mental.

Let’s distinguish two kinds of psychological modules. Call one type F-modules in honor of the sorts of modular systems that Jerry Fodor’s initial modularity hypothesis was meant to depict. In The Modularity of Mind, the seminal contemporary work on modularity, Jerry Fodor (1983) lays out the properties characteristic of modular systems such as reflexes, face recognition, and the five senses. The five senses have most or all of the following features: They are domain specific, processing very narrow types of input (eyes, light; ears, sounds), mandatory, involve limited central access, are fast, informationally encapsulated, produce shallow outputs, operate on fixed neural architecture, are open to characteristic and specific breakdown, and demonstrate a characteristic pace and sequencing in ontogeny. Input systems differ from central systems in that central systems are slower and take input in a format that allows computation or deliberation over representations in a non-modally specific format, in principle, over everything we know.19 So we say that thinking and problem solving can integrate, inter alia, information received from the senses plus from previous thinking. These differences lead Fodor to introduce “Fodor’s First Law of the Nonexistence of Cognitive Science”: “the more global (e.g., the more isotropic) a cognitive process is, the less anybody understands it. Very global processes, like analogical reasoning, aren’t understood at all” (p. 107).

Some philosophers and mind scientists object to F-modules, but not it seems because they really doubt there are F-modules, or what is different, doubt that conceiving of certain systems modularly can be useful, but because they believe incorrectly that accepting that there are F-modules requires accepting much stronger functional and anatomical autonomy claims than they in fact require (Churchland, Koch, & Sejnowski, 1993).

Call another type of module E-modules, in honor of Paul Ekman’s work on basic affect programs, as, for example, subserve the seven universal facial expressions (Darwin, 1871; Ekman, Levinson, & Friesen, 1985; Flanagan, 2000, 2003, 2009). We call a module an E-module if it has all the properties of F-modules and in addition it necessarily involves powerful affective and conative aspects, where the feeling and action tendency have features of automaticity, rapidity, and cognitive impenetrability. So, for example, the feeling of alarm and compassion will be experienced when the child is falling into the well. The action to save the child can be stopped, but only by conscious deliberation/veto from central systems.

E-modules may—indeed they seem to—require input from F-modules, since one needs to see the child falling into the well to experience “alarm and compassion.” But even if modules are sometimes layered on top of modules, E-modules are worth distinguishing from F-modules upon which they may depend, because they necessarily carry heavy affective components, which the five senses or reflexes associated with the ANS, qua modules, do not. Some F-modules, reflexes, for example, instigate action tendencies as do E-modules but often or typically (consider the knee jerk or pupil contractions or expansions) do so without any emotional involvement. E-modules are affectively loaded and always motivate (but do not always necessarily produce) a bodily response.20 Indeed, every important list of basic moral attitudes (Confucius, Mencius, Aristotle, Hume, Darwin, Ekman, Strawson) depicts them as having powerful emotional oomph that is not associated necessarily with F-modules. Of course, moral modules, such as compassion, involve cognitive appraisal: “this is a child falling into a well.” But, once the child falling into the well is perceived, the emotional reaction is automatic, fast-on-its-heels. Even King Herod, knowing that the child is a first-born male Jew, cannot immediately override his impulse to want to save the child. The feeling of distress and the impulse to save the child will happen, and this feeling and the action-tendency are what Pylyshyn calls “cognitively impenetrable.” Of course, King Herod can, as he did, think it is a good idea, all things considered, to kill all the first-born males. But doing so, especially if the killing is up close and personal, will involve overriding powerful impulses not to do so.

6. Darwin and moral modularity

  1. Top of page
  2. Abstract
  3. 1. Introduction: Moral modularity
  4. 2. Ancient modularity
  5. 3. Twenty-first century modularity
  6. 4. Classical and contemporary modularity
  7. 5. Modules with and without emotional oomph
  8. 6. Darwin and moral modularity
  9. 7. Normativity
  10. 8. Normativity and modularity
  11. References

Was Darwin (1871) himself a modularist when it comes to explaining moral sense? Maybe. Here is what he says in the Descent in his most explicit gloss on human moral sense:

In order that primeval men, or the ape-like progenitors of man, should become social…they must have acquired the same instinctive feelings…They would have felt uneasy when separated from their comrades, for whom they would have felt some degree of love, they would have warned each other of danger, and have given mutual aid in attack or defence. All this implies some degree of sympathy, fidelity, and courage….[T]o the instinct of sympathy…it is primarily due that we habitually bestow both praises and blame on others, whilst we love the former and dread the latter when applied to ourselves; and this instinct no doubt was originally acquired, like all the other social instincts, through natural selection…. [W]ith increased experience and reason, man perceives the more remote consequences of his actions, and the self-regarding virtues, such as temperance, chastity, &c., which during earlier times are…utterly disregarded come to be highly esteemed or even held sacred…Ultimately our moral sense or conscience becomes a highly complex sentiment–originating in the social instincts, largely guided by the approbation of our fellow-men, ruled by reason, self-interest, and in later times by deep religious feelings, and confirmed by instruction and habit. (pp. 498−500)

The fact that Darwin uses “instinct” is promising for reading him as a modularist. But the fact that he adds that with “increased experience and reason, man perceives the more remote consequences of his actions” indicates that the “highly complex sentiment” that emerges in actual worlds is a complex, partly cognitively and historically conditioned competence. This seems plausible, indeed it seems obvious. No sensible student of ethics could think otherwise. But if true, it has one possibly worrisome implication for the cognitive science of morality. Fodor’s First Law, recall, claims that we can expect success in the study of a given faculty to the degree that it is modular, be the modules anatomical ones, F-modules, or E-modules. The more standard modular properties that a faculty, system, or subsystem has and the greater degree to which the faculty realizes these properties, the more we can expect to learn about the faculty. Central systems have few of these properties and to a small degree, and thus are likely to remain intractable by psychology and cognitive science.21 At least, this is true if we accept that progress in studying the moral mind depends on the degree to which it is modular. How modular is it? No one knows yet. Then again, one could plausibly conclude that if the explanatory value of cognitive science ends where the story of moral modularity ends, then social psychology, sociology, and anthropology are nicely poised to take over the explanatory work. Suppose we continue to gain more and more clarity and depth in understanding and explaining human moral psychology as well as the structure and texture of moral life across the earth, do we also deepen our understanding of what is really good? The psychobiological idea of moral modularity is perhaps best expressed in quotes as the “moral modularity” hypothesis, where the moral modules refer to the equipment in human first nature that can be grown, moderated, modified, and/or suppressed (usually some of each) to become what a particular community deems a good person. What is the relation between being deemed a “good person” and being a good person without the quote marks? Or, to put the matter another way, assuming there are “moral modules” what is the right way to tune them so that they become or undergird true virtue, so that they support the moral personalities of genuinely good people? This is the problem of normativity.

7. Normativity

  1. Top of page
  2. Abstract
  3. 1. Introduction: Moral modularity
  4. 2. Ancient modularity
  5. 3. Twenty-first century modularity
  6. 4. Classical and contemporary modularity
  7. 5. Modules with and without emotional oomph
  8. 6. Darwin and moral modularity
  9. 7. Normativity
  10. 8. Normativity and modularity
  11. References

David Hume, himself a great naturalist who saw the basis for all legitimate moralities in human nature and human social life, taught us that there is a problem about “oughts,” specifically that you cannot derive “oughts” from “is’s” (some add that for similar reasons you cannot derive values from facts). After the famous passages in which Hume depicts the fallacy, he continues making normative judgments about moral means and ends for well over 100 more pages. Was Hume inconsistent? Not at all. Hume’s wildly misunderstood point is that one ought (rationally) beware people, especially religious types, who claim to have demonstrated that it is morally necessary that you ought to ϕ because it is the moral law (or divine law). What Hume established is that one (rationally) ought not claim to have derived a (moral) “ought” from any set of facts, even allegedly transcendentally true ones. This is because “oughts” are one among the innumerable kinds of things that cannot be derived = demonstrated = deduced from propositions that do not also contain words like “ought.” That you ought to eat breakfast does not follow from any set of facts about you and nutrition. Is it a good idea to eat breakfast? Of course. Can you derive the conclusion that you ought to eat a healthy breakfast from facts about nutrition, plus facts about your own desires for health and well-being, plus those of your loved ones? No. Minimally, you will need to add a premise to the effect that one ought to do what achieves one’s ends. And then the problem allegedly repeats. How does one justify this “ought”? Is it derived from facts, and, if not, where does it come from? The usual move here is to say that such prudential “oughts” are not really so mysterious, but that moral “oughts” are. This is a very hard case to make. If one accepts that health is a reasonable end, then perhaps eating well is a means. But how do we justify the end of health itself? Answer: People agree that it is a good idea. This makes many nervous since extended to ethical ends, and not just means (bring on the angels with their heavenly trumpets), this might lead us to think that ethical ends are specified and justified by considerations of human nature, human desire, human consensus, and nothing else. But this is disturbing only if one antecedently thought ethics could have some deeper transcendental grounding, which the naturalist denies. This denial of transcendental foundations is the naturalists’ response to a 2,500-year long research program in the West that has been utterly unprogressive. Rather that naturalists point to the existence, ever since Darwin, of a credible alternative explanation for the why, how, and wherefore of human morality.

Most of the important things, the things that matter, cannot be proved (=demonstrated). Truths of mathematics and pure logic can be; the rest, not. Most true things do not follow and cannot be derived from the true things that warrant their assertion. For example, it did not follow from the facts that we wanted to cross the Hudson River between New York and New Jersey, and that the George Washington Bridge would do the job, that we ought to build the George Washington Bridge. We could have swum, boated, built any number of other style bridges, or given up the aspiration to cross the Hudson altogether. Nor does it follow from the fact that every person so far has died that you will die and that you ought to believe this is so. That is, it does not follow that you are mortal despite the fact that all persons who have ever lived before you died. It does not follow logically that you ought to seek to promote the happiness of everyone because you want to be happy and so does everyone else. The first (the George Washington Bridge) was a terrific idea and has worked out nicely. The second (your mortality) is something that, if you are rational (not delusional), you ought to believe in. The third, in the form of the golden rule, is widely thought to be a good idea even if it is neither a self-evident truth of a system nor a theorem of a system with self-evident axioms.

We can change the philosophical dialectic some if we ask: What is it about normativity that is supposed to make naturalists worry or feel impotent? The answer is that there need be no such worry. Most inference is inductive and abductive (inference to the best explanation among available ones), not deductive. When people say “this follows from that” they do not mean normally that it deductively follows. Or if they do, they should not. So when Watson says to Sherlock Holmes “brilliant deduction!” he is always referring to an abduction (read C.S. Peirce [1955] for more on this ubiquitous logical form). The dialectic, really the rhetoric, about “is” and “ought” might make one think that a naturalist is not allowed to speak or argue about norms or better and worse practices and that his resources only allow description and causal explanation. City-planners, architects, and bridge-builders are naturalists. They engage in discussions within communities of fellow planners (with differential power); the communities then generate ends (from among sets of “good ideas”) and create structures that achieve these collectively generated ends. So too with morality.

One odd feature about the aura that is alleged to surround “oughts” is that it might make one think that nonnaturalism is equipped to explain how “oughts” are derived. But it is not. “Oughts” are not (normally) derived or deduced. But “oughts” do not sit out in thin air in such a way that only the ghost-whisperers can explain. “Oughts” are reasoned to in a holistic network that operates over both propositions about facts and propositions about antecedently settled values and “oughts,” all of which are open to conversational challenges. Neither the ends nor the means are deduced.

Hume’s fallacy is avoided if you do not claim derivation or deduction or demonstration. Claim that the ends you wish to defend—happiness, fulfillment, flourishing, hedonistic sensuality, material wealth, top-dog dominance, or obedience to a God—are reasonable, and that such and so are means to these ends, and then we can talk, debate, and see whether your reasons hold up.22 Morality is a nonmysterious public domain (although the mystifiers are a dime a dozen among theists and deontologists), and everyone is equally entitled to recommend good ideas for sociomoral practice. Ethics is like engineering. Specify ends. Talk about whether the ends involve good, worthy goals. Specify means, evaluate them in terms of effectiveness, costs, and so on. Then do the right thing. Everything, means and ends, is open to conversational challenge.23

8. Normativity and modularity

  1. Top of page
  2. Abstract
  3. 1. Introduction: Moral modularity
  4. 2. Ancient modularity
  5. 3. Twenty-first century modularity
  6. 4. Classical and contemporary modularity
  7. 5. Modules with and without emotional oomph
  8. 6. Darwin and moral modularity
  9. 7. Normativity
  10. 8. Normativity and modularity
  11. References

If normative discourse is not restricted by commitment to naturalism, and it is not, then what is to be said if anything about the normative implications of SIM? That is, supposing SIM is true, are there any normatively useful insights to be extracted? In closing, we examine two possibilities.

8.1. The virtue of tolerance

One normative consequence that SIM theorists suggest follows (now in the nondeductive sense) from SIM has to do with the warrant for greater tolerance. By recognizing the role played by each of the five modules in morality, but at the same time recognizing the indeterminate range of each module and the different degrees of emotional tuning (the strength of feelings of “intolerance, indignation, and disgust” that different people have to different practices, as Lord Patrick Devlin put it, when speaking about reactions to male homosexuality in the United Kingdom circa 1959), we can understand more deeply why moral tolerance is rational, why tolerance is a virtue, even if it is not particularly well-enabled by any of the original modules taken singly. Thus, one lesson to extract from the research discussed earlier about American Liberals and Conservatives, which found that Liberals tend to focus on the sprouts of fairness/reciprocity and harm/care, whereas Conservatives tend to appeal to all five of the sprouts and thus moralize manners in a way Liberals do not, might be that we be tolerant of these differences in moralizing. Haidt and Graham (2007, p. 113) draw something like this conclusion: “[r]ecognizing these... foundations as moral (instead of amoral, or immoral, or just plain stupid) can open up a door in the wall that separates liberals and conservatives when they try to discuss moral issues.” The idea is that recognition of our shared moral psychology in the five modules warrants greater understanding of where others are coming from, how they could moralize as they do, and so on.

Now there is an interesting and important question here as to whether the tolerance warranted is (a) to involve tolerance in the sense of understanding where the other is coming from (and how it was that her judgment is as good as mine). If one thinks that morality is wholly determined by the modules, whatever they are, in interaction with a form of life, whatever it is (and why ever it is what it is), then one might be driven to think that both the psychological genealogy and the substance of different valuations are to be respected and tolerated. If, however, one thinks that nonmodular central systems get a say because the function of morality involves adjustments and adaptations to new situations and new worlds and that certain factual issues are at stake when people disagree about sociomoral practices, then not all views need to be tolerated understandable that the innate modules could have been grown her way rather than my way, and that we both feel strongly about the matter at hand—that’s what Darwinian (E-)modules are designed to do, to provide powerful affective oomph to our sociomoral reactions in their domain), or (b) to involve accepting as good, certainly not as equally good.

8.2. Which sprouts to grow, which modules to activate, and how much?

We return for one final word on a question that has persisted throughout this paper and needs to be brought into full view in closing. What force, what significance, do any facts about basic human nature have with regard to the question of how we ought to construct ourselves and our sociomoral worlds? At the start we set out two versions of MMH:

Moral ModularityDescriptive. Human nature contains sprouts for four (or five) different moral sprouts or modules.

Moral ModularityNormative. Moral excellence involves growing all four (or five) sprouts to maturity.

The descriptive thesis we have argued is credible. SIM (with the possible addition of some role for central processing), in particular, is an empirically plausible modularity hypothesis.

But the normative thesis faces a host of problems. Insofar as our human nature has sprouts in it, there are good sprouts and bad sprouts, possibly these are one and the same. The histories of ethics and moral psychology from their beginning in Confucius, Buddha, and Aristotle to more recent theorizing are filled with botanical conjectures; however, the overwhelming consensus is that, depending on the demands of the environment and the histories of a people, different sprouts in our natures will be grown and cultivated in different ways. At the same time other sprouts in our nature will be conceived of as weeds or poisons (the “poisons” in Buddhism are also sprouts or modules—lust, anger, delusion). Weeds, of course, are simply plants we do not like. Mencius is pretty much the only philosopher, classical or contemporary (except for Frans de Waal in some Pollyanna-ish moments) who speaks only of good sprouts (Flanagan, 2009). Xunzi, another classical Chinese philosopher, is famous for challenging Mencius by claiming that “People are bad.” Xunzi, like Mencius, Buddha, and SIM theorists are looking at what they take to be the sprouts. So how many sprouts are there—if indeed sprouts are the right metaphor? How exactly are the sprouts to be individuated? Are the modules in SIM fine grained enough? The modules in SIM—each module—arguably contain more than one sprout (a better analogy may be that each module is a sprout that can be grown and pruned in ways so as to develop certain possible branches, as in the ploctopus case)—in which case there are ploctopus-like possibilities at both the level of each module and at the level of the whole integrated system. This would make the model very complicated, but then the domain may require such complexity. The ingroup/loyalty module, we take it, contains the sprouts (or limb buds) for loyalty, patriotism, territoriality, suspicion, and resentment. When are such things, such fruits, blossoms, or limb buds, good or bad, virtues or vices? The right answer seems that it depends. It depends on a host of factors outside, or in addition to, the modules, on features of human history, the current environment, and so on. Morality is an accommodation to interpersonal life in social worlds that are not the same as the worlds in which the original equipment evolved.

The question remains, what to do? Use your head. Pay attention to what you, we, or they sensibly aim to accomplish and how to accomplish said aim, keep deadly conflict to a minimum, and keep your eyes resolutely on what your innate nature wants, at times, to do and why. The modules inside you, being Darwinian, will make you feel cocky, assured that you see things correctly. Watch out for this; it can lead you astray. Generating cocky feelings (righteousness) is how they are designed; it is what the modules are supposed to do in the adaptationhistorical sense. But we live in different worlds now than then, so keep your eye on whether your moral confidence really has warrant in this world, in the one you have actually living in now. Use those advanced mental abilities that go beyond the modules, that involve wisdom (zhi), phronesis, or good old common sense to track ways of living well that suit you for achieving what Aristotle called eudaimonia (=flourishing).

This recommendation—to use your head (all of it, not just the modules)—is of course an old idea. But it nicely reconnects the cognitive science of morality with normative ethics in a way that involves a clearer understanding as to why we humans have so much trouble agreeing on normativity. The reason has nothing to do with any inherently mysterious features of the logic of “oughts” or with the metaphysics of values or norms. It has to do with the fact that our natures as persons are not fully specified by our biological natures. The existentialists had it right. We humans, in virtue of being social, cultured, and very smart, are co-creators with “Mother Nature” of our being. Our most unusual biological feature, and an evolutionary advantage so far, is the enormous plasticity we possess to make and remake ourselves in ever-new ways, to grow and train our innate sprouts to fit our current condition. There are a plethora of realistic options to “grow” and “do” human nature. It is not surprising that we are still getting used to the fact that the right answer to the question “how shall I live” is not given by our human nature nor it is unequivocally there to be read off the external world. Once we accept fully that we are Darwinian creatures, material beings living in a material world, not living in a world where morality has transcendental sources, it might be easier to accept that the questions of how we individually or collectively ought to live is not one of those where there is the one, right answer. Neither knowledge about the innate modules nor deeper wisdom about general-purpose cognition will reveal how we ought to be and to live. But knowledge of both, especially knowledge of the modules, can teach a great deal about the initial settings, various natural trajectories, and the limits of the possibility space for making and remaking ourselves. And knowledge is power.24

Footnotes
  • 1

     This proposal can be found in the work of Haidt and his colleagues (Haidt, 2001; Haidt & Graham, 2007; Haidt & Joseph, 2004, 2007) as well as Fiske (1991, 1992, 2004), Greene (2003), Hauser (2006), Shweder (1990), and Shweder and Haidt (1993). In developing their five-foundation (module) system, Haidt and colleagues were greatly influenced by and expanded on the work of Shweder and Fiske. In this paper we focus on the work of Haidt and his colleagues (the five-foundation model) for two reasons. First, this five-foundation model is presented as (largely) compatible with Shweder’s three-foundation model (the “ethics” of autonomy, community, and divinity) and Fiske’s four-foundation relational model (communal sharing, authority ranking, equality matching, and market pricing) (Graham et al., 2010). Second, Graham et al. (2010) claims that the five-foundation model outperforms models with fewer foundations with regard to both fit and parsimony. While we do not take a stand on these claims of mapping and model fit, we do take Haidt’s five-module system to be representative of the field. Caveat: Mark Hauser (2006) has a modularity model that consists of one faculty, which may or may not be compatible with (or be able to house) social intuitionist modularity (SIM) modules (as submodules). We are unopinionated—officially agnostic—about Hauser’s model in this paper.

  • 2

     This is John McDowell’s useful terminology for discussing Aristotle. Aristotle makes few surmises about first nature, but he is all on about the question of how persons are transformed from first nature, whatever it is, to mature, rational, and virtuous second nature.

  • 3

     Despite logical concerns about deriving “oughts” from “is’s,” empirical scientists have tried to cross the gap. In the early 1970s, Lawrence Kohlberg claimed to possess an empirical theory of moral psychology that enabled us (finally) to derive “ought” from “is,” to plot moral development, and to establish the philosophical adequacy of (his and Kant’s) the highest stage of moral development (Kohlberg, 1971). Kohlberg’s idea failed (Flanagan, 1991), but the renewed interest in empirical moral psychology in the last 20 years has been accompanied by revisiting the “is-ought” barrier with an eye for gaining some normative consequences from the study of moral psychology, even if these normative consequences involve only ruling out certain rule-based theories such as consequentialism or Kantianism as psychologically unrealistic rather than resolving which among the number of psychologically realistic ways of configuring moral personality is best (Flanagan, 1991, 2002; Greene, 2003; Hauser, 2006).

  • 4

     “In general, having these four sprouts within oneself, if one knows how to fill them out…If one can merely fill them out, they will be sufficient to care for all within the Four Seas. If one merely fails to fill them out, they will be insufficient to serve one’s parents” (Mencius, 2A6).

  • 5

     If having all four virtues well developed is not “normal” in the sense of “usual” (and it is not for Mencius, who is nostalgic for a past Golden Age when virtue was normal) in the way having four limbs is, then we are owed an explanation of how and why the current environment fails to pull for the development of the four cardinal virtues in the same way we would need an explanation for odd numbers of limbs.

  • 6

     The two senses of adaptation, original-historical and current ecology, are the favored ways among philosophers and biologists of making the distinction (plus or minus some) that psychologists make in terms of proper and actual domains. Haidt and Joseph (2007) follow Sperber in dividing module triggers into a proper and actual domain: “Sperber (1994) refers to the set of objects that a module was “designed” to detect as the proper domain for that module. He contrasts the proper domain with the actual domain, which is the set of all objects that in fact trigger the module” (p. 16). The language we use of adaptationhistorical and adaptationcurrent ecology is preferred in philosophy, inter alia, it marks the possibility (which is actual) that even if the proper and actual triggers are the same, the environment may have changed so dramatically that this is no longer functional. Then and now I want to kill you if you steal my sexual partner (same proper and actual triggers). But the technological extensions of myself now available, guns for example, make the impulse more destructive now than it was when the ice melted at the end of the Pleistocene. At that time all I could do was chase you away or try to exact revenge with my fists or found objects. Now, I am more likely to succeed in killing you. Adaptation talk emphasizes phenotypic traits while domain talk emphasizes the set of circumstances that trigger these traits. What we (and Sperber, Haidt, etc.) are really interested in is the relationship between phenotypic traits and their domains; we merely choose to emphasize the former.

  • 7

     A number of researchers (Casebeer, 2003; Casebeer & Churchland, 2003; Churchland, 1996; Flanagan, 1991, 1996; Haidt, 2001; Johnson, 1993) take virtue theory to be better supported by, or more consilient with, the empirical findings of psychology and neuroscience than are competing ethical theories. One of the main points of support drawn from the research is that the declarative rules of deontology or consequentialism are often either not appealed to or are appealed to as mere post hoc rationalizations of prior intuitions (Haidt, 2001).

  • 8

     In correspondence, Haidt says he is not committed to five modules being the right number. There might be more.

  • 9

     In correspondence, Haidt rejects the stronger normative reading of SIM. While Haidt does not believe all the sprouts should be nurtured, he claims Harm and Fairness are grown universally to some degree and are necessary for a minimally decent society.

  • 10

     If it is found that a given moral modularity model cannot account for some aspect of morality, the model might be able to proliferate modules to account for the relevant aspect. The SIM itself expands on the (largely) compatible moral taxonomies that suggest fewer foundations (Kohlberg, Gilligan, Fiske, Shweder, etc.). The real trouble comes if the relevant aspect cannot be accounted for in terms of modules (see below on central systems).

  • 11

     While Haidt and his colleagues regularly call these five foundations “modules,”Haidt and Joseph (2007) make a somewhat different, possibly weaker claim: “[a]ll we insist upon is that the moral mind is partially structured in advance of experience so that five (or more) classes of social concerns are likely to become moralized during development” (p. 381).

  • 12

     The surface and deep structure language is familiar from Chomsky. Marc Hauser (2006), an advocate of Chomskyean views, has a modularity view that is different from MMM and SIM and postulates a single moral faculty akin to the universal grammar (UG) claimed to be embedded in the language acquisition device (LAD). We say nothing directly about such a model here.

  • 13

     Charles Goodman, a wise comparative philosopher, writes to us: “Mencius very clearly reasons in accordance with ingroup/outgroup although he has no special sprout for it.” See 3A5: “Does Yi Tzu truly believe... that a man loves his brother’s son no more than his neighbor’s new-born babe?” See also 4B29: “Now if a fellow-lodger is involved in a fight, it is right for you to rush to his aid with your hair hanging down and your cap untied. But it would be misguided to do so if it were only a fellow-villager. There is nothing wrong with bolting your door.” Perhaps Mencius does not have a separate sprout for ingroup/loyalty because he thinks it is relatively easy to develop the relevant attitudes. It seems more likely that he sees an appropriate degree of partiality for those close to you as included in some or all of the other sprouts. This point is stated, perhaps even more strongly than Mencius’s other views should justify, at 4A27: “The content of benevolence is the serving of one’s parents; the content of dutifulness is obedience to one’s older brothers; the content of wisdom is to understand these two and to hold fast to them; the content of the rites is the regulation and adornment of them; the content of music is the joy that comes of delighting in them.”

  • 14

     In communication, Haidt claims filial piety is concerned with authority not ingroup/loyalty, and while the ingroup/loyalty and authority/respect modules tend to co-occur, they have different evolutionary histories and underwrite different adaptations.

  • 15

     It is worth noting that some who hold the mind to be massively modular—that is, the mind is modular as the rule, not the exception—would object to the contrast we present between central systems (reason, wisdom, etc.) and modular systems. While we do not speak directly to the massive modularists, we stand by our claim that moral competence seems to require some nonmodular elements (for more, see fn. 21 below on Cosmides and Tooby).

  • 16

    Van Norden (2007, p. 352) suggests that the four Mencian sprouts govern four parts of life, where, for example, propriety (li), would not even match Haidt’s authority/respect since it is all about beauty. On this interpretation, benevolence (ren) would be the only good match between MMM and SIM.

  • 17

     An important question we sidestep concerns “realism” versus “instrumentalism”—does the fact that studying a system, the body or the mind, modularly is profitable mean that there are modules? See Flanagan and Feldman (“Is the Moral Modularity Hypothesis Testable?” unpublished data) where this issue is engaged directly.

  • 18

     Richard Lewontin gives an excellent example from population genetics. A breed of fruit fly may display a normal distribution of the number of belly hairs at a certain temperature, but change the temperature and the distribution of belly hairs on each fruit fly’s belly changes. What number of belly hairs is trying to express itself? This question has no answer for individuals or the population independent of the environment.

  • 19

     Fodor is not trying nor does he believe it is possible to give a definition of “modules” in terms of necessary and sufficient conditions. Although the nine characteristics of modules would seem to give us clear-cut ways to distinguish between modular input systems and general purpose central systems, the case is not so clear-cut. The problem here is two-fold. First, Fodor’s conception of modules allows for degrees with regard to the above listed nine properties of input systems. Some will be more encapsulated than others both informationally and morphologically (mutatis mutandis for the rest). Second, input systems need not have all of the nine properties, but all such systems will have most of these properties. Rather, the sine qua non of input systems, and what prevents central processes from counting as modular, is the informational encapsulation of the system. Throughout the work, Fodor points out that we have made a great deal of progress in understanding input systems (sense modalities, reflexes, and language mechanisms) but very little progress in our attempts to understand more general cognitive processes.

  • 20

    Flanagan (2000, 2007) calls the basic emotions “proto-moral.” This prefix “proto“may be useful to describe the moral modules as well because they are “moral” originally only in the sense that they can become part of moral character when they are moralized inside a community and worked over, possibly grown as in the case of compassion (harm/care), possibly suppressed or moderated in the case of ingroup/loyalty.

  • 21

     Leda Cosmides and John Tooby have written widely on the issue of modularity. For them, thinking in terms of adaptive modules is a necessary part of situating psychology with the natural sciences (specifically biology) and removing it from the road-blocking intuitions of what they call the standard social science model (SSSM). While their work has far too much depth and breadth to treat fully here, Cosmides and Tooby (1992) is instructive of their general arguments and the differences between their conception of modularity and the more modest modularity considered in this paper. Much like the SIM theorists, Cosmides and Tooby take various empirical findings (e.g., the fact that people perform better on Wason selection tests that involve a component of social exchange than those that do not) to support the claim that certain aspects of the mind can profitably be modeled modularly; these modules are then activated and tuned by one’s ecological situations to form an “evoked culture” (p. 218). However, Cosmides and Tooby take odds with the Fodorian distinction between modular and central systems (which they claim to be a core tenet of SSSM) that we have been pulling on throughout this paper: “Even vigorous advocates of modularity have held so-called central processes, such as reasoning, to be general-purpose and content-independent (e.g., Fodor, 1983)” (p. 165). But they claim their findings “support the contrary contention that human reasoning is governed by a diverse collection of evolved mechanism, many of which are specialized, domain-specific, content-imbued, and content-imparting” (pp. 220−221). This seems to put Cosmides and Tooby at odds with our contention that moral competence requires some central processing via reasoning or wisdom. Three things need to be said here. First, Cosmides and Tooby could be taking an instrumentalist stance such that (and now in-line with Fodor) by treating “central systems” modularly we will be able to make more progress in the investigation of such systems, regardless of whether these systems are made of modules. Second, the real problem with the central/modular system contrast may be caused by conceiving of the two types of systems as qualitatively different. Fodor emphasizes that differences between various systems are ones of degree; those systems that demonstrate more of his nine criteria and to a greater degree are more modular than those that show less. Thus, the conclusion to take from Cosmides and Tooby may be that reasoning and other canonically central systems are more modular than previously thought. Third, while the differences between central and modular systems may be largely quantitative, the sine qua non of modular systems is their information encapsulation, and it seems clear that our mind cannot be informationally encapsulated all the way up. Thus, while it is largely an empirical issue as to whether it may be worth investigating canonically central systems modularly, whether the differences between central and modular systems have been exaggerated, and whether the mind is more modular than the modest modularist would like to admit, we stick to our contention that moral competence requires some central (non-informationally encapsulated) processing in order to train up the modules, check them against one’s current environment for appropriateness, break ties, etc.

  • 22

     See MacIntyre (1959) and Searle (1964) for two classics on Hume on “is-ought.” In addition to Flanagan’s arguments (Flanagan, 1991, 1996, 2006; Flanagan, Sarkissian, & Wong, 2007), others who argue along similar lines are Casebeer (2003) and Mark Johnson (1993).

  • 23

     Alan Gibbard (1990) and Simon Blackburn (1998) develop a theory they call “norm expressivism,” which explains normative discourse as involving “it makes sense _____” judgments. Such judgments are open to conversational challenge. Nothing is deduced. The view is usually described as a kind of noncognitivism, which is misleading since “it makes sense to _____” judgments involve beliefs about world-norm pairs.

  • 24

     Of course, gaining this knowledge (even of the modules) requires central processing.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction: Moral modularity
  4. 2. Ancient modularity
  5. 3. Twenty-first century modularity
  6. 4. Classical and contemporary modularity
  7. 5. Modules with and without emotional oomph
  8. 6. Darwin and moral modularity
  9. 7. Normativity
  10. 8. Normativity and modularity
  11. References
  • Blackburn, S. (1998). Ruling passions. Oxford, England: Clarendon Press.
  • Brosnan, S. F., & De Waal, F. B. M. (2003). Monkeys reject unequal pay. Nature, 425, 297299.
  • Brown, D. E. (1991). Human universals. Philadelphia: Temple University Press.
  • Casebeer, W. D. (2003). Natural ethical facts: Evolution, connectionism, and moral cognition. Cambridge, MA: MIT Press.
  • Casebeer, W. D., & Churchland, P. S. (2003). The neural mechanisms of moral cognition: A multiple-aspect approach to moral judgment and decision-making. Biology and Philosophy, 18, 169194.
  • Churchland, P. M. (1996). The engine of reason, the seat of the soul: A philosophical journey into the brain. Cambridge, MA: MIT Press.
  • Churchland, P. S., Koch, C., & Sejnowski, T. J. (1993). What is computational neuroscience? In E. L.Schwartz (Ed.), Computational neuroscience (pp. 4655). Cambridge, MA: MIT Press.
  • Cosmides, L., & Tooby, J. (1992). Cognitive adaptations for social exchange. In J.Barkow, L.Cosmides, & J.Tooby (Eds.), The adapted mind: Evolutionary psychology and generation of culture (pp. 163228). New York: Oxford University Press.
  • Darwin, C. (1871). The descent of man and selection is relation to sex. New York: D. Appleton and Associates.
  • Ekman, P., Levinson, R. W., & Friesen, W. V. (1985). Automatic nervous-system activity distinguishes among emotions. Science, 221, 12081210.
  • Fiske, A. P. (1991). Structures of social life. New York: Free Press.
  • Fiske, A. P. (1992). Four elementary forms of sociality: Framework for a unified theory of social relations. Psychological Review, 99, 689723.
  • Fiske, A. P. (2004). Relational Models Theory 2.0. In N.Haslam (Ed.), Relational models theory: A contemporary overview (pp. 325). Mahwah, NJ: Lawrence Erlbaum.
  • Flanagan, O. (1991). Varieties of moral personality: Ethics and psychological realism. Cambridge, MA: Harvard University Press.
  • Flanagan, O. (1996). Ethics naturalized: Ethics as human ecology. In L.May, M.Friedman, & A.Clark (Eds.), Minds and morals: Essays on ethics and cognitive science (pp. 1944). Cambridge, MA: MIT Press.
  • Flanagan, O. (2000). Destructive emotions. Consciousness and Emotion, 1(2), 259281.
  • Flanagan, O. (2002). The problem of the soul: Two visions of mind and how to reconcile them. New York: Basic Books.
  • Flanagan, O. (2003). Ethical expressions: Why moralists scowl, frown and smile. Cambridge companion to Darwin (pp. 377398). Cambridge, England: Cambridge University Press.
  • Flanagan, O. (2006). Varieties of naturalism. In P.Clayton & Z.Simpson (Eds.), The Oxford handbook of religion and science (pp. 430452). Oxford, England: Oxford University Press.
  • Flanagan, O. (2007). The really hard problem: Meaning in a material world. Cambridge: MIT Press.
  • Flanagan, O. (2009). Ethical expressions: Why moralists scowl, frown and smile (revised). Cambridge companion to Darwin (pp. 413434). Cambridge, England: Cambridge University Press.
  • Flanagan, O., Sarkissian, H., & Wong, D. (2007). Naturalizing ethics. In W.Sinnott-Armstrong (Ed.), Moral psychology (vol. 1): The evolution of morality: Adaptations and innateness (pp. 125). Cambridge, MA: MIT Press.
  • Fodor, J. (1983). Modularity of mind. Cambridge, MA: MIT Press.
  • Gardner, H. (1983). Frames of mind: The theory of multiple intelligences. New York: Basic Books.
  • Gardner, H. (1993). Multiple intelligences. New York: Basic Books.
  • Gardner, H. (2006). Multiple intelligences: New horizons. New York: Basic Books.
  • Gibbard, A. (1990). Wise choices, apt feelings. Cambridge, MA: Harvard University Press.
  • Graham, J., Nosek, B. A., Haidt, J., Iyer, R., Koleva, S., & Ditto, P. H. (2010). Broadening and Mapping the Moral Domain: Development and Validation of the Moral Foundations Questionnaire. Manuscript submitted for publication.
  • Greene, J. (2003). From neural ‘is’ to moral ‘ought’: What are the moral implications of neuroscientific moral psychology? Nature Review of Neuroscience, 4, 846849.
  • Haidt, J. (2001). The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review, 108(1), 814834.
  • Haidt, J., & Graham, J. (2007). When morality opposes justice: Conservatives have moral intuitions that Liberals may not recognize. Social Justice Research, 20(1), 98116.
  • Haidt, J., & Joseph, C. (2004). Intuitive ethics: How innately prepared intuitions generate culturally variable virtues. Daedalus, 133(4), 5566.
  • Haidt, J., & Joseph, C. (2007). The moral mind: How five sets of innate intuitions guide the development of many culture-specific virtues, and perhaps even modules. In P.Carruthers, S.Lawrence, & S.Stich (Eds.), The innate mind (Vol. 3) (pp. 367392). Oxford, England: Oxford University Press.
  • Hauser, M. (2006). Moral minds: How nature designed our universal sense of right and wrong. New York: HarperCollins.
  • Johnson, M. (2003). Moral imagination: Implications of cognitive science for ethics. Chicago: University of Chicago Press.
  • Kohlberg, L. (1971). From is to ought: How to commit the naturalistic fallacy and get away with it in the study of moral development. In T.Mischel (Ed.), Cognitive development and epistemology (pp. 151235). New York: Academic Press.
  • MacIntyre, A. C. (1959). Hume on “Is” and “Ought”. The Philosophical Review, 68(4), 451468.
  • Mencius. (2008). Mengzi:With selections from traditional commentaries. Bryan W. Van Norden (Trans.). Indianapolis, IN: Hacket Publishing.
  • Peirce, C. S. (1955). Abduction and induction. In J.Buchler (Ed.), The philosophical writings of Peirce (pp. 150156). New York: Dover.
  • Schwartz, S. H., & Bilsky, W. (1990). Toward a theory of the universal content and structure of values: Extensions and cross-cultural replications. Journal of Personality and Social Psychology, 58, 878891.
  • Searle, J. (1964). How to Derive “Ought” from “Is”. The Philosophical Review, 73(1), 4358.
  • Shweder, R. A. (1990). Cultural psychology: What is it? In J. W.Stigler, R. A.Shweder, & G.Herdt (Eds.), Cultural psychology: Essays on comparative human development (pp. 143). New York: Cambridge University Press.
  • Shweder, R. A., & Haidt, J. (1993). The future of moral psychology: Truth, intuition, and the pluralist way. Psychological Science, 4, 360365.
    Direct Link:
  • Shweder, R. A., Much, N. C., Mahapatra, M., & Park, L. (1997). The “big three” of morality (autonomy, community, and divinity), and the “big three” explanations of suffering. In A.Brandt & P.Rozin (Eds.), Morality and health (pp. 119169). New York: Routledge.
  • Sperber, D. (1994). The modularity of thought and the epidemiology of representations. In L. A.Hirschfield & S. A.Gelman (Eds.), Mapping the mind: Domain specificity in cognition and culture (pp. 3967). Cambridge: Cambridge University Press.
  • Van Norden, B. W. (2007). Virtue ethics and consequentialism in early Chinese philosophy. Cambridge, England: Cambridge University Press.
  • De Waal, F. (1991). The chimpanzee’s sense of social regularity and its relation to the human sense of justice. American Behavioral Scientist, 34, 335349.
  • De Waal, F. (1996). Good natured: The origins of right and wrong in humans and other animals. Cambridge, MA: Harvard University Press.