THE SPILLOVER THESIS
The “Spillover effect” denotes the tendency for norms of autonomous public reasoning to seep into and possibly undermine reasonable religious ways of life. When a boy goes to public school in the United States, he might be asked to develop his critical reasoning skills. Even (or perhaps especially) outside of social issues classes, he will be asked to think critically using his own considered “autonomous” judgment. In literature classes, science classes, or even history classes, the quality of his work will be assessed based wholly or in part on his ability to critically assess evidence. Officially at least, faith, revelation, and religiously based restraints on conduct receive little or no reinforcement. Religion is tolerated and if conflicts occur, accommodations are reached when interference with the school's public mission is negligible. The boy may be exempted from some activities, at the price of being separated from his peers.
In class, the historical figures, scientists, and fictional characters the boy studies are very likely to be praised chiefly for their exercises of autonomous judgment and seldom for their piety or deference to moral and epistemic authorities. He will read about the steadfast Howard Roark from the novel The Fountainhead who used his mind to build amazing new structures despite social pressure to conform coming from all sides, or Galileo Galilei who spent the last eight years of his life under house arrest for publishing his claim that the Earth revolved around the sun, or Thomas Jefferson and other founding fathers who insisted on revolution against a world founded on the divine right of kings. There are no elegies to the wisdom of Abraham who trusted God's wisdom so blindly that he was willing to kill his own child on mere faith. With autonomy and critical reason so highly prized, what chance is there that this boy might still defer to religious authority on matters of private moral conscience? Won't autonomous reason simply become ingrained as a reflex, seeping over into all parts of life, even the parts where “ethical subservience”Footnote 1 is officially tolerated? This is the fear among politically liberal citizens of faith. The question is a factual one that has until now only been addressed on an intuitive level by political philosophers. We shall return to it anon.
FACTS AND NORMS
Facts matter when evaluating the plausibility of our political theories. For example, one of the criticisms leveled at liberalism in general and John Rawls' “Theory of Justice” in particular is that it is psychologically implausible for people to value conceptions of the good life that prioritize faith in revealed truth and at the same time adhere to a conception of justice that demands an exercise of public reason when justifying coercion. The theory demands too much. It is not logically inconsistent, the criticism goes, but its effect is to impose a psychological toll above and beyond what can be reasonably required of the average religious citizen. A different way of stating the problem, one made by Steven Macedo, is that there are “cultural spillovers” from liberal public norms that generate systems of unequal psychological taxation, with citizens of faith typically "paying at a higher rate" (Macedo Reference Macedo2000). It is unclear whether this particular appeal to equality carries much weight, since no liberal needs to be committed to an idea of equal psychological effect on a society's citizens.Footnote 2 Nonetheless, the idea that there is something wrong with making it especially burdensome for reasonable people to live together under a liberal regime seems like a critique worth taking seriously. If it turns out that liberalism requires things that most people are incapable of or even only things that are rather difficult to do, then we might have reason to think that it is not a good theory around which to organize political life.Footnote 3 This does not mean that human beings could never live this way or that some people are not capable of practicing more advanced forms of political living together, but, if politics is not about perfectionism but about respecting individual ways of living, then this respect requires that ordinary citizens be able to obey the law without excessive difficulty.Footnote 4 Politics may set higher aspirational standards, which citizens might be praised or otherwise rewarded for meeting, but the minimum required ought to be within the ordinary citizen's grasp.
We do not want to say that it is always wrong to require something that is very difficult or impossible. That would go too far. Citizens who find themselves on the tail end of the relevant distributions of material, physical, psychological, or spiritual attributes may unavoidably find themselves in a higher “psychological tax bracket.” We may still impose requirements on pedophiles or psychopaths who have difficulty exercising dominion over their impulses even as we must concede that it is very unlikely or even impossible that they will succeed. It is also permissible not to build our political world exclusively around a concern for the weakest and the most vulnerable among us, or for that matter those who are so spiritually needy that any contact with the liberal polity is conceived of as profane. The legal standards for product liability need not take for a reference point what the most intellectually dull consumer might think about a product's proper use. We can safely leave for another time a discussion of what can be required of those who are very different from the average citizen.Footnote 5 More obviously, extraordinary circumstances may require extraordinary moral resolve. But rules for ordinary life and rules for extraordinary circumstances need not be the same.
While difficulty is not necessarily a knock-down reason to reject moral theories in the abstract, difficulty carries extra weight when assessing expressly political theories and ways of living together. In particular, it is weighty when the political theory is a liberal one, claiming to accommodate the kinds of conceptions of the good life likely to result from conditions of freedom. If the purpose of a liberal theory about living together is to regulate interactions among people who only share a minimal moral conceptionFootnote 6 (Larmore Reference Larmore1996), then broad accessibility ought to be a cardinal virtue.Footnote 7
RAWLS AND POLITICAL LIBERALISM
In writing Political Liberalism, one of John Rawls' aims was to make liberalism more acceptable to individuals who have moral reasons to endorse liberal political institutions, but may not practice a very liberal way of life. He does this by rejecting the idea that endorsing liberal political institutions needs to be grounded in a conception of the person that prizes individual moral autonomy above all else (Rawls Reference Rawls1993). No longer would it be necessary for citizens to affirm the basic principles of justice upon which political coercion is based for the reasons traditionally given by liberal philosophers. The principles can now be affirmed for different reasons, and in particular for religious reasons. The idea of the “burdens of judgment” is particularly important for this move. The idea stems from the realization that there are many citizens who affirm religious doctrines (some of them quite stringent), but also accept that reasonable people living under desirable conditions of freedom and equality are bound to end up disagreeing about what it means to live a good life. While some or nearly all of these views about the good life may be mistaken, the task of living together requires tolerance and a certain amount of humility with respect to the kinds of reasons that can justify political coercion. All religious doctrines that can affirm the need for consent to justify temporal coercion are potential candidates for an overlapping consensus: a set of expressly political principles for the purpose of living together and getting along. Because individuals find moral reasons to endorse these principles from within their comprehensive conceptions of the good life, these political principles are not just a modus vivendi, but a partly free-standing political morality. Those doctrines, religious or otherwise, that reject the need for this form of justification are then termed unreasonable and reasons adduced from such doctrines are given no weight in considering public policies.
In Political Liberalism, Rawls affirms that moral autonomy is necessary only for political purposes. Civic education is therefore required to teach autonomy solely for the purposes of political entreaty. If a comprehensive conception of the good life affirms political liberalism alongside traditional gender roles or divisions of labor, this is no longer a target for government rectification through education (Freeman Reference Freeman2007, p. 239), except perhaps inasmuch as it undermines a sense of justice (Okin Reference Okin1991). It should be added that for Rawls, some autonomy in private affairs is still a prerequisite for qualifying as a reasonable comprehensive conception of the good life. A belief that individuals have no choice about their religious commitments would fail the test of the reasonable. Belief in private autonomy grounds the principle of toleration. Members of Rawls' overlapping consensus must value everyone's capacity for a conception of the good and this includes the freedom to change one's religion.Footnote 8 To that extent, civic education still has something to say about private autonomy.
However, to say that citizens must value private autonomy to some degree is not to say that they must make it the central focus of their moral commitments. What matters is that individuals have some freedom of choice, some domains of autonomy, not that this autonomy be the most extensive available. There is a wide possible range of views about what counts as sufficient individual autonomy, many of which fall far short of the license found in parts of American mainstream culture. When it comes to the actual kinds of religious creeds operating within the United States, many probably do accept the burdens of judgment even as they would sometimes place severe restrictions on the otherwise legally permissible choices of their members. And presumably few support laws forbidding religious conversion.
The evidence from middle-class America seems to support this view. Alan Wolfe interviewed a representative sample of 200 middle-class Americans to find out if the apparent revival of evangelical religion translated into support for increased use of religious justifications for public policy. Even in Tulsa, Oklahoma, the heartland of conservative Protestantism, respondents were usually unwilling to say that others should be made to follow religious beliefs whose absolute truth the respondents endorsed. Wolfe calls this middle-class consensus, belief in the “modest virtues.” Americans believe there is a right and a wrong way to act, but also believe one ought to refrain from imposing one's views on others through coercion or excessive proselytizing. When polled about people of religious faiths other than their own, most believed they were capable of being moral people. Many also expressed a desire to withhold judgment, consistent with modesty in how they view their religious faith. They did not like the politicization of their faith, in great measure because of their disdain for politics. While Wolfe found that most middle-class Americans took their religion seriously, very few seemed to believe it should be the most important guide for establishing rules about how other people should live (Wolfe Reference Wolfe1999). Religion is too important to be forced on people.
If we are worried about autonomy, perhaps it is not members of orthodox religions we should primarily worry about. As Jeffrey Spinner-Halev points out, orthodox religionists in liberal societies rarely live the way they do because of a lack of alternatives, or even a lack of considered reflection. Many kinds of religious lives are very much chosen. To be a religious conservative in liberal society be it Jewish, Christian, or Muslim is to constantly be confronted with and reflect upon the differences between the life one leads and the lives everyone else leads (see also Eberle Reference Eberle2002). To remain different under these circumstances is frequently a life more “chosen” than the life of someone who unreflectively adopts the views of the dominant culture (in effect, the ones dominant in public schools). In their everyday lives also, orthodox citizens of faith often exercise meaningful autonomy. Even if some foods are proscribed, there is still a wide menu available. Even if some professions are considered taboo, occupational choices are still usually available. Even in areas where important choices are constrained, like marriage, sons and daughters still often have input into and choices between a more restricted set of alternatives.Footnote 9
Even if it is morally acceptable for conservative religious citizens to endorse a robust conception of political autonomy and at the same time deny autonomy a central role in their conceptions of the good life, there is further question about how sustainable this compromise will prove to be. Even if Rawls the philosopher has cleared conceptual space for his view, it remains to be seen whether it is realizable in practice.
IS CIVIC EDUCATION TOO BURDENSOME?
In the liberal debates about civic education that followed the publication of Political Liberalism, Amy Gutmann argued that it is implausible to think that one could educate students to be autonomous and free for the purposes of political decision-making, and then expect them to give up that acquired skill when dealing with more personal and private matters: “It is probably impossible to teach children the skills and virtues of democratic citizenship in a diverse society without at the same time teaching them many of the virtues and skills of individuality and autonomy. We say many because individuality and autonomy place limits on the extent to which they can be taught. Both entail living one's life according to one's own best lights because one judges this a good way to live” (Gutmann Reference Gutmann1995, p. 563). Eamonn Callan concurs, arguing that comprehensive doctrines and public reason tend to effectively compete for the minds of citizens:
This is a high-wire act. One falls off if allegiance to the favored comprehensive doctrine influences (perhaps unconsciously) political judgment, and precisely because one is sure of its truth, the influence may be hard to resist. One also falls if acceptance of the burdens of judgment eats away at the convictions that public reason pushes outside its boundaries. (Callan Reference Callan1996, p. 19)
This kind of reasoning leads Callan and Gutmann to similar conclusions. Gutmann argues that because of this inconsistency, one ought to drop the pretense of trying to accommodate reasonable views about the good life that reject the thesis that individual freedom and autonomy should be the sole or primary lens through which the non-public world should be evaluated. Religious students may not be exempted from social studies classes or civics classes on the grounds that their freedom to exercise their religion is being unduly infringed by classes that will teach civics skills in a way that might undermine their religious beliefs. Nor may class curricula be changed, if this means failing to teach citizens important critical-analysis skills. If it comes to a choice between inculcating the skills necessary to navigate a liberal democracy and undermining potentially reasonable ways of life, Gutmann's choice is clear. Similarly, Callan argues that liberal legitimacy requires the inculcation of autonomy as well as a sense of justice. A government is not legitimate unless it is chosen by citizens operating from a more or less autonomous standpoint. Since autonomy is necessary for legitimacy as well, accommodation of religious citizens is again overridden.
THE HUMAN MIND: E PLURIBUS UNUM
Research in psychology, however, casts doubt on these intuitions about the relationship between citizenship and autonomy and this in turn saves many people from having to choose between their religious beliefs and a liberal public life. The case made by Guttman (1995), Callan (Reference Callan1996), and Macedo (Reference Macedo2000), which seems to be implicitly accepted by McConnell (Reference McConnell, Macedo and Tamir2002), Spinner-Halev (Reference Spinner-Halev2000), Tomasi (Reference Tomasi, Macedo and Tamir2002), and others, is that education that seeks to produce individual autonomy for the purposes of citizenship inevitably has spillover effects into the way an individual conceives of his or her individual moral commitments. In the following section, we argue that these spillovers are far from inevitable because the human mind is not structured the way most of us intuitively believe it to be. Instead, we are led to believe that robust programs of civic education can but need not lead to students casting a more critical eye on their private moral commitments. The tension between religion and liberal civic education is therefore overstated. Four distinct areas of research help buttress this claim.
The Fundamental Attribution Error
Evidence from the last 30 years of psychological research is overwhelming. There are systematic ways in which our everyday intuitive view of the human mind is flawed. Character is not a stable set of related traits and the reason we believe it is has to do with something called the Fundamental Attribution Error (FAE). When it comes to interpreting the behavior of others, individuals usually overestimate the importance of a fundamental character trait, and underestimate the importance of situation and context. When analyzing civic education, political philosophers like all of us are tempted to attribute a single logically coherent personality to children, one that would naturally use the information or modes of thinking acquired in a civics class to analyze other parts of their lives, including what is preached in Sunday school. The FAE is well documented and stems from research in attribution theory. Coined by Lee Ross, the observation stems from a seminal study done by Edward Jones and Victor Harris (Reference Jones and Harris1967), confirmed by Jones and Nisbett (Reference Jones, Nisbett, Jones, Kanouse, Kelley, Nisbett, Valins and Weiner1972), and Ross (Reference Ross and Berkowitz1977). Other names for the same phenomenon are the “over attribution effect” or the “correspondence bias.” By 1982, when David Watson published a well-respected literature review, he felt that it was not unreasonable to conclude that the hypothesis was now “firmly established” (Watson Reference Watson1982, p. 698). The FAE helps explain why political theorists and casual observers who have a theory of mind based on the idea of a unitary personality might be unreasonably suspicious about spillovers. If you assume that each individual has a set of identifiable and consistent traits across all contexts, the spillover thesis is very plausible. Any change in a person's “master” value set would trickle down or “spillover” into contexts outside that in which the new value was internalized. To be sure, the existence of the FAE does not yet show that we are disposed to compartmentalize. The FAE only shows that we tend to misperceive the reasons for which other people behave, but it does not tell us anything about how their minds actually work.
Evolutionary Psychology
If human beings are not behaving for the reasons for which we think they are behaving, then “why are they doing what they're doing”? The last two decades have seen a shift in how many, if not most, neurologists and psychologists view the human mind. MIT cognitive neuroscientist Steven Pinker documents this evolution in his book How the Mind Works. The previous standard model relied on a nature/nurture distinction where nature set capacities or limits and culture provided the substance. Pioneered by anthropologist John Tooby and psychologist Leda Cosmides, a new and increasingly accepted paradigm is that the human mind is a series of mental modules that operate certain kinds of innate programs. This updated model does not just posit that behavior is a product of interactions between nature and nurture, but actually side steps this debate. Instead, the mind is thought to have a series of computational faculties engineered by natural selection (indeed, it is sometimes called the “computational model” of the mind). Far from being a blank slate or even imbued with “instincts,” the mind has evolved to process information in very specific ways. The mind is not infinitely plastic within a set of natural limits, but instead is more rigid and less unitary than previously thought.
Cosmides and Tooby (Reference Cosmides, Tooby, Hirschfeld and Gelman1994) argue for the idea of domain-specific cognitive functioning. Evolutionary psychologists hold that domain specificity, also known as modularity of the mind, is a survival mechanism that humans developed millions of years ago during the Pleistocene period for the purpose of problem-solving in various contexts. Since evolution is slow, the brains of contemporary humans are wired in much the same way as their early ancestors. Specific domains are processed innately by humans through the development of modules as determined by natural selection. Those considered relatively innate include folk biology (knowledge of plants and animals), the capacity for language, intuitive psychology (theory of mind and social skills), and practical physics (gravity and basic mechanics) (Atran Reference Atran2005).
Modules other than the above core ones are developed over time based on cultural and environmental requirements, thanks to the adaptability of the relatively large human brain (Fodor Reference Fodor1998). Religion and politics may be classified as such culture-bound modules. Cosmides and Tooby (Reference Cosmides, Tooby, Hirschfeld and Gelman1994) would never argue that there are different modules for public and private moral reasoning since this distinction was likely inapplicable during 99 percent of human history where the brain's evolution occurred. However, there is significant evidence that the mind is well trained to recognize in-group and out-group members and adjust behavior accordingly. Even more impressively, there is some evidence that individuals are able to make inferences about distinct types of groups based on very little information, suggesting some kind of pre-existing ideal types of groups (Lickel, Hamilton, and Sherman Reference Lickel, Hamilton and Sherman2001). Lickel, Hamilton and Sherman find that individuals quite consistently perceive clusters of different kinds of ideal type groups: intimacy groups (friends, family, spouses), task groups (teams, offices, juries), social categories (women, African-Americans), and loose associations (people who like techno music, people waiting at the bus stop). This “intuitive taxonomy” also helps shape a moral framework of appropriate and inappropriate interaction. Rules as to how to behave with family and friends, fellow church goers, and other citizens belong to very different group models of interaction (as hypothesized by Deutsch (Reference Deutsch1975) and illustrated by Fiske (Reference Fiske1991)).
Lickel, Hamilton, and Sherman (Reference Lickel, Hamilton and Sherman2001) argue that specific features tend to account for a group's classification into one or another of the lay types they identify. These features, they argue, serve as a trigger for classification and a set of associated norms. “Perceivers intuitively comprehend interactions among group members in terms of certain relational rules by which those interactions are governed” (Lickel, Hamilton, and Sherman Reference Lickel, Hamilton and Sherman2001, p. 133). If true, then we have evidence for compartmentalization of moral judgment, since the mind appears to make intuitive distinctions between the kinds of rules that operate among different types of groups.
These ideal types just discussed also have implications for how the brain organizes information. For example, Sherman, Castelli, and Hamilton (Reference Sherman, Castelli and Hamilton2001) show that this group typology is spontaneously and implicitly used when encoding social information. Using a variation on the recognition memory paradigm developed by Taylor et al. (Reference Taylor, Fiske, Etcoff and Ruderman1978), they found that individuals manage information in their memory according to the type of group that is being perceived. Should political morality and one's comprehensive conception of the good life be conceived of as rules for different kinds of groups, what we learn from one may not easily pass to the other. The Sherman study does not suggest that people completely compartmentalize knowledge, only that memory is selectively associative. Students may have to be prodded more than one thinks for modes of thinking acquired for one group purpose to be applied to another. Even if the connection is made, the individual may yet consciously reject the new rule as inapplicable in a different context.
Psychology cannot give us good moral reasons to legitimize different rules for different groups of people, but it can tell us something important about what is relatively easy and relatively harder for the human brain to do. When relative ease of practice matters morally (as it does for political liberalism), then we should be aware of what the mind does intuitively, versus what it requires conditioning to do. In this case, we now know that individuals intuitively have room for at least four different kinds of rules of engagement based on context.
Multiple Intelligences
Showing that the human mind operates according to separate and distinct modules (or programs, to use the computer metaphor) helps clear the air and helps us better understand why people think and behave differently according to context. Different social settings will trigger different forms of thinking and behavior. Depending on the setting, different programs will “boot up” and condition our behavior. A related view is Howard Gardner's multiple intelligences theory. First introduced in 1983, it presents the idea that people have cognitive competence in disparate areas of functioning that may have little bearing on the other areas; these numerous intelligences include linguistic, musical, logical-mathematical, spatial, body-kinesthetic, interpersonal, and intrapersonal (with naturalistic, spiritual, and existential as later “candidate” additions) (Gardner Reference Gardner1983). Gardner believed the reason they exist is that people have multiple neural modules with their own memory banks and ways of operating, demonstrable by brain injury to an individual area that hinders one ability but does not alter another. (Unlike Cosmides and Tooby, he does not base this claim on evolutionary psychology.) Multiple intelligences suggests that not only do different people not think alike, but the same person does not always think the same way depending on the nature of the object of his attention.
Gardner did not suggest that we have different mental modules for thinking about private and public morality, but his research does suggest that there are different modules available for any given task. Multiple intelligences and areas of competence that are at least partially independent suggest that learning does not trickle or flow through the brain as easily as one might imagine. If we extend the mind-as-computer metaphor, the mind appears to be made up of many smaller computers of disparate architecture and different processing speeds, with their own memory banks that are linked, but not completely so. Upgrades to one computer through learning does not guarantee performance improvements elsewhere. Spillovers are possible, but not obviously so, particularly in light of what we know about how intuitively there is space for different norms to govern relations between individuals and members of different group types.
Domain Specificity
There is also evidence of cultural evolution that provides some support for the idea that people exercise domain-specific types of reasoning across different contexts. Atran (Reference Atran2001) explains that human cultures support stable and rapid transmission of ideas and memes that help solve relevant and recurring environmental problems. These are easily remembered and processed by the human brain, and facilitate the retention and understanding of useful or important ideas that are hard to learn, such as science or religion. Modularity eases the ability of humans to orient themselves while managing these ideas, minimizing confusion and maximizing efficiency. For example, while one may value belief in deities and mystical ideas, such cognitive activities may not necessarily be helpful when applied to a given problem-solving activity, and thus they are not given pride of place at times when they are less relevant. This does not mean they lose relevance for the person; it merely states that faith will not be directly useful for solving engineering problems. In other words, individuals activate different (culturally specific) mental modules depending on the context in which they find themselves. These modules piggy-back on biologically evolved ones.
Atran's empirical studies using evolutionary psychology's concept of domain specificity demonstrate the existence of a natural human ability to think differently across situations. His 1998 study on Lowland Maya Amerindians and Midwestern Americans — two culturally distinct populations — shows an innate categorization process used to organize types of knowledge in the specific domain of folk biology (Atran Reference Atran1998). Although some general processing ability is necessary to first form domain-specific processing, modes of thinking are quite different from one domain to the next, and categories are initially formed so that these differing modes may be assigned accordingly. Atran likens folk biology to Kant's “synthetic a priori knowledge” and defines domain specificity as “an innately determined cognitive structure embodying the naturally selected ontological commitments of humans” (Atran Reference Atran1998, p. 554–555). What he says here is that people are naturally wired to be able to think in various ways — categorically and specifically — when dealing with problems or issues of ongoing importance to their lifestyle. They just need to fill in the specifics. A good analogy is language. Individuals are hardwired for linguistic capability, but not for any specific language. We are not suggesting that individuals are hardwired for politics, only that we have a hardwired ability to compartmentalize based on context. This would apply as much for politics as it does to baseball, churches, rock concerts, or court houses. How people separate or define the boundaries of each context is a cultural and social process (Turiel Reference Turiel1983).Footnote 10 This is where the kind of civic education one receives plays a potentially important role.
In particular, domains of appropriateness appear to play a strong role in conditioning our behavior. Like trains, we can be switched from one track to another, but once on one track, we must wait for another switch before moving to a different track. Very early we learn to differentiate domains of appropriateness: home, school, church, sports, and so on. Seminal experiments by Hugh Hartshorne and Mark May (Reference Hartshorne, May and Munsinger1971) showed that cheating among school children was highly context dependent. Some children cheat at home but not at school, or vice versa. Some cheat at some kinds of tests but not others. Moreover, children tend to exhibit consistencies in the contexts where they cheat, not in whether they cheat or not. Very few children were always honest or dishonest, but almost all children were consistent in the types of situations in which they cheated. More recent work by Prencipe and Helwig (Reference Prencipe and Helwig2002) confirms the notion that children understand different types of moral values and differentiate moral judgments based on context. Other experiments done by Philip Zimbardo at Stanford showed that individuals (adults this time) put into high stress situations (like a fictitious prison) tend to lose their habitual predispositions in often dramatic ways (Haney, Banks, and Zimbardo Reference Haney, Banks and Zimbardo1973). Context dictates behavior far more powerfully than is commonly assumed by more unitary theories of personality. Most importantly, Matsumoto, Guerin, Hartshorne and May, and Zimbardo show that moral behavior and judgment vary across contexts. For example, for some children it was acceptable to cheat on one kind of test rather and not others, and for most study participants it was acceptable to verbally abuse or even physically abuse people in prisons but not outside. This is particularly relevant to our argument since it demonstrates an ability to compartmentalize moral judgment.
The concept of domain specificity helps explain these phenomena. Atran states that a range of stimuli is preferentially processed per each domain (Atran Reference Atran2001; also Lord, Ross, and Lepper Reference Lord, Ross and Lepper1979). The specific combination of sensory information coming in, as filtered by the mind for relevance, determines the particular way of thinking that is activated, meaning that various places and situations have their own accompanying modes of thought. For example, being in a classroom would activate a style of thinking different from that used in driving a car, talking to someone personally close, or attending a religious ceremony. Predictably, this type of mode activation has implications for a person's thoughts, behaviors, judgments, and decision-making processes. It is important to note that both cultural constructions and naturalistic phenomena determine the sensory information available for processing. Reviewing this literature leads psychologist Elliot Turiel to conclude: “The findings of this body of research demonstrate not only that behaviors vary by context, but also that people's orientations to a given situation are more complex than that expected from dichotomous categories like individualism and collectivism or independence and interdependence” (Turiel Reference Turiel1996, p. 83).
Culture and ecology play large roles in determining the types of activities and situations to which a person is exposed as well as the settings in which these exposures occur. And culture influences the way its people think in various domains because it determines what is important and deserving of attention. Necessarily, then, living in a particular culture and geographical area influences the types of category-based cognitive processing modes that inhabitants develop for use in both private and public spheres. Individual differences mix with these influences to determine a person's thoughts and behaviors in situations, since some people may be more affected by external influences than others (Ross, Bierbrauer, and Hoffman Reference Ross, Bierbrauer and Hoffman1976; Turiel Reference Turiel1996).
We have four distinct areas of research that suggest that it is possible for a person to maintain beliefs related to personal philosophies or spiritual belief systems all the while detaching from them in his or her thinking processes while the person functions in realms other than the philosophical or spiritual (e.g., political, social, etc.). This is likely a learnable (and almost ubiquitously practiced) skill that employs existing cognitive structures evolved to do similar things. While there is some worry about generalizability and ethnocentrism with any small-N study of human psychology, these findings appear relatively robust. Turiel and Wainryb (Reference Wainryb and Turiel1994) find that this kind of contextual differentiation in moral judgment occurs in both hierarchical and more egalitarian settings. Spiro (Reference Spiro1993) finds related evidence of multiple and conflicting norms reflecting concepts of personal entitlements and a “cultural ideology of interdependence” in places as varied as Bali, India, Nepal, Pakistan, Japan, and Burma. Although many individuals no doubt allow their religious beliefs to influence their stances on various political issues (and vice-versa), doing so is not neurologically compulsory; in fact, conclusions drawn from the study of anthropology, evolutionary psychology, and moral and developmental psychology might suggest that doing the opposite — maintaining boundaries between domains of mental functioning — is not only possible, but also very common.
AUTONOMOUS AND HETERONOMOUS
The empirical research and theories summarized in this article suggest that it is possible for children and adults to be autonomous for the purposes of citizenship and partly heteronomous in their everyday lives. Just like any other context, political context matters as to how someone is going to think about an issue. This evidence goes against what some have treated as common-sense linkages between religious and political beliefs. For example, Kenneth Greenawalt claims, “For almost everyone, the ways to determine religious truth and political truth are significantly linked. One linkage involves ‘negative checks:’ What we confidently believe about ordinary facts and ethics serves as a check on what we will accept as religious truth and as sound political philosophy” (Greenawalt Reference Greenawalt1995, p. 122). The truth may be rather more complicated. We have to have particular reasons or “triggers” for running these negative checks. If the moral rules examined are part of the same domain, it is more likely that we will run these checks. If we have been trained to run checks across domains, then it is also more likely that we will do so. Otherwise, however, we may not.
On further reflection, it does not seem at all odd that the profit maximizing entrepreneur can also be a philanthropist, or that politicians can uphold high standards of public ethics while leading deplorable private lives. We are confronted with these phenomena all the time and end up probably more puzzled than we should be. If we feel we have good reason to distinguish the contexts, or if we have been conditioned to do so, nothing will seem odd about faithfully following the teachings of a cleric in one's private affairs and using one's own best judgment in public matters. Pointing out the inconsistency would seem strange and beside the point. Of course, the reverse is also true. If we have been told repeatedly growing up that moral judgment and political judgment are part of the same domain of moral reasoning, then spillovers will indeed probably occur.
Eamonn Callan argues that the burdens of judgment and other central politically liberal notions “impose educational requirements that bring autonomy through the back door of political liberalism” (Callan Reference Callan1996, p. 21). If the research into domain specificity is correct, then this back door has a lock if we choose to use it. Individuals are able to compartmentalize various domains of cognition and it comes more naturally than many assume. The extent of the spillover effect will depend on the way social norms separate various contexts and the way individuals are socialized to think about politics and the good life. The “default mode” seems, however, to be differentiation. This is not consistent with the “inevitable” spillovers envisioned by many of those who have thought about liberal civic education and its effect on religious commitment.
The Fundamental Attribution Error will always tempt us to see one context as subordinate to the other. The entrepreneur is in fact really a cutthroat capitalist, and only gives to charity because of the tax deductions, or public pressure, or because his wife makes him do it. The politician is really a selfish hypocritical liar, and his public record only shows the pressure of ethics laws, his political party, or public scrutiny. In reality, we know that honesty, virtue, and other character traits are more or less in evidence depending on the context and that a man who nudges his golf ball out of a bad lie is not necessarily going to cheat on his wife or cheat his business partner (or even more likely to do so than someone who plays the game fairly).Footnote 11 An average person is only rarely honest across the board, kind across the board, non-violent across the board, and so on. Indeed, we are likely to wonder about the sanity of the people who appear to be just that. Given the sometimes extreme compartmentalization of behavior depending on context that we exhibit, it may not at all be the case that the exercise of political autonomy for a normally subservient housewife in a fundamentalist Christian sect is a tragic farce, as most secular humanists are disposed to believe.Footnote 12 She may have a robust sense of political equality and yet accept a relatively inegalitarian home life as long as she is capable of accepting that other women ought not to be coerced into inegalitarian family arrangements. It might well be unlikely, if she does not receive sufficient civic education to establish a clear notion of the boundaries between public and private, but it is not an idea that can be dismissed a priori. This example is not intended to make light of serious and legitimate questions about influence and oppression within the family. It instead encourages us to abandon a reductionist view of the individual (and the fundamentalist Christian housewife) that fails to account for neurological complexity.
Susan Moller Okin argued that liberal philosophy has been insufficiently skeptical of the role family relations play in the moral development of children, and as such ought to be more carefully scrutinized by principles of justice as fairness (Okin Reference Okin1991). If we want citizens to be capable of making more just political decisions, we ought to be concerned with family relations that were once thought to be outside the domain of conventional justice. For Okin and others, the family is one of the primary breeding grounds for a sense of justice and if oppression and injustice pervade our families, it is difficult to see how children can grow up with a well-formed sense of justice. If research into domain specificity is correct, some — though not all — of Okin's concerns can be allayed. While it is true that a sense of justice will in part be developed in the family, this sense of justice does not have to be tainted by spillovers from the demands of a less egalitarian family life. As long as a robust distinction between comprehensive conceptions of the good life and political values can be maintained by both family and school, then very different kinds of values and processes of decision making may come to dominate in each. Okin might respond that inegalitarian family structures are themselves unjust, but if they are, it is not primarily because they inevitably turn children into bad citizens. After all, we can safely assume that many good citizens were raised in overly patriarchal family arrangements. The fact that they tend to reproduce unjust family structures does not mean they necessarily reproduce unjust political institutions.Footnote 13
Critical thinking, gullibility, faith, skepticism, and other kinds of attitudinal dispositions are not an automatic or constant feature of anyone's personality. Triggered by specific cues, autonomy is a variable that is at least partially independent from other mental processes. The evidence we bring to bear suggests that human beings possess different mental modules that can kick in based on various cues, particularly physical or social context. There is still some disagreement in the literature about the degree to which the domains are independent from one another. Connectionist models (as opposed to the mostly computational ones we have discussed this far) will posit that the mind is made up of interconnected neural networks. Information produced in one node of the network (say a mental module) may yet serve as an input to another through channels that have been created over time. This makes it impossible to state that compartmentalization is airtight. Information can flow across modules and the manner in which it does is primarily learned. Fortunately, our argument does not depend on compartmentalization being airtight.Footnote 14 It is certainly possible for individuals to take knowledge or ways of thinking and apply them to other domains. This is a central feature of learning. However, we believe some spaces can be carved out, or at the very least, the barriers to entry for some domains are much higher than others. This article only depends on two claims that the connectionist models do not invalidate: We have some brain architecture that enables us to think differently across different contexts and this extends to moral and ethical judgments. Spillovers can happen, even spontaneously, but they are not as inevitable or ubiquitous as is assumed by the conventional theory of mind adopted by political theorists who have studied the subject.
CONCLUSION
There is a tradition in political philosophy dating back to classical times that suggests that virtue makes a person fit to rule. The evidence from psychology suggests that the practice of virtue is in great measure domain specific. Perhaps it is desirable on some grounds that those who rule should also set the example in other domains of their lives, but when it comes to choosing leaders, private virtue may not be a reliable epistemic guide. It does not appear that democracy depends on the private virtue of its citizens either. Within some range, a terrible father or an immoral clergyman may still make a good voter just as a selfless mother and a reputable business man may vote poorly. If autonomy is essential to the practice of public virtue, then the evidence suggests that one can have a healthy dose of it in public affairs and somewhat less of it in other contexts.Footnote 15 We tend to run the critical thinking module only in certain contexts and not so much in others.Footnote 16 One can certainly criticize some or all compartmentalization on moral grounds, but in doing so we should have an appreciation for the difficulty of changing the way the mind is structured. Compartmentalization is neurologically easier than many philosophers believe. Perhaps there is something unsettling about the fact that individuals are less consistent across contexts than they appear. Perhaps we should encourage more spillovers in moral learning from one area of the brain to another. These are conversations for another day. Spillovers from liberal civic education are not inevitable byproducts of political liberalism. They have to be promoted and nurtured in order to take root. This should take the conversation away from the pursuit of autonomy writ large and towards the manner in which autonomy is encouraged or discouraged in a variety of more disaggregated contexts.
