As the religious juggernaut of the modern West, America’s unusual piety has confounded longstanding assumptions about religion and modernity and contributed to a reappraisal of traditional secularization theory [Stark and Finke Reference Stark and Finke2000]. Yet while scholars have puzzled over the unusual religiosity of the American people, they have found another outstanding attribute of the United States—its exceptionally secular political institutions—relatively unremarkable. In comparative terms, America’s institutional secularity is striking; according to one metric, the United States features the most absolute separation of religion and state of any nation in the world [Fox Reference Fox2008].
Scholars typically point to America’s high degree of religious diversity and its strong constitutional language as factors contributing to its strict “separation of church and state” [Bruce Reference Bruce2011; Fox Reference Fox2008; Gill Reference Gill2008; Martin Reference Martin1978]. But these familiar explanations are less convincing than they first appear, because America is not exceptional in either regard. Australia, for instance, is also highly religiously diverse, lacks a state church, and enjoys nearly identical constitutional protections for religion. Yet unlike the United States, Australian political institutions are far more amenable to religious influences [Monsma and Soper Reference Monsma and Soper1997], despite Australians’ somewhat lower rates of religious belief and practice [Norris and Inglehart Reference Norris and Inglehart2004]. In education, for example, Australia not only permits religious instruction in public schools, but lavishly funds a burgeoning denominational school sector [Maddox Reference Maddox2014].
How should we account for these arrangements? The question is of more than idle interest. Scholars have shown that the way religion is or is not incorporated into political institutions can have important religious, political, and economic consequences [Barro and McCleary Reference Barro and McCleary2003; Casanova Reference Casanova1994; Gill Reference Gill2008; Martin Reference Martin1978; Stark and Finke Reference Stark and Finke2000]. Theoretically, meanwhile, the question speaks directly to the ongoing reevaluation of the venerable “secularization thesis.” Once dominant in the social sciences, the classical prediction that religion would inevitably decline as societies became more modern [Berger Reference Berger1969; Weber Reference Weber, Gerth and Wright Mills1946] has fallen on hard times in recent years. Increasingly, scholars acknowledge that secularization can lead to a diverse array of “secular settlements”, or relatively stable sets of policies governing the role of religion in public life [Gorski and Altinordu Reference Gorski and Altinordu2008]. Indeed, a flurry of recent works has begun to explore the remarkable variety of secular settlements visible in the modern world [Berger, Davie, and Fokkas Reference Berger, Davie and Fokas2008; Casanova Reference Casanova2006; Kuru Reference Kuru2009; Martin Reference Martin2005; Monsma and Soper Reference Monsma and Soper1997]. In this context, America’s strikingly secular political institutions—and Australia’s relatively less secular ones—should be understood as but two possible secular settlements among many, whose origins are neither self-evident nor inevitable, and whose characteristics have consequences for a range of social outcomes.
In this paper, I juxtapose the cases of Australia and the United States to improve our understanding of how and why secularization can lead to such different secular settlements, even among otherwise similar countries. In particular, I examine the fate of religious instruction in each nation’s public schools from the mid-nineteenth to the mid-twentieth centuries. It was during this period that today’s secular settlements began to crystallize, with devotional instruction declining in American schools but holding fast in Australia. I explain these divergent patterns by focusing on a neglected factor in the existing literature: the state. Apart from a general consensus that established churches tend to lead to more “assertive” forms of secularism [Kuru Reference Kuru2009; see also Gorski Reference Gorski and Dillon2003; Martin Reference Martin1978; Reference Martin2005], the state—as an organization with an institutional structure and capacity for autonomous action—has been largely overlooked as a potential independent force shaping the development of modern secularity. This is so despite awareness that the structure of the state has profoundly shaped politics and policies in multiple other fields [Fetzer and Soper Reference Fetzer and Soper2005; Halfmann Reference Halfmann2011; Smith Reference Smith2008; Weir and Skocpol Reference Weir, Skocpol, Evans, Rueschemeyer and Skocpol1985].
I argue that America’s decentralized system of educational administration facilitated the more rapid and extensive secularization of American education. America’s permeable state structure provided multiple access points for religious minorities seeking to challenge majoritarian religious practices from below, and a range of supports for educational elites seeking to introduce professionalizing reforms from above. By contrast, Australia’s insulated state structure centralized curricular decision-making, making secularizing reforms more difficult for religious minorities and educators alike. The state thus had both mediating and constitutive effects on secularization. Both by structuring political contests over religious policy, and by creating (or failing to create) structural conditions and incentives that helped to generate actors with secularizing interests, the structure of the state helps explain the secular settlements we see today.
The paper proceeds as follows: I begin by laying out the contrasting trajectories of religious instruction in Australian and American public schools from the mid-nineteenth to the mid-twentieth centuries, and discuss the different administrative systems that coordinated education in each country. I then trace how these systems structured political conflicts among religious groups, and how they shaped the development of education as a profession in each country, showing in each case how administrative structures inhibited secularization in Australia, while promoting it in the United States. I conclude by considering how this comparison helps us to rethink our approach to secularization, and by theorizing the role that the state plays in secularization more generally.
Secular settlements and the state
While much scholarship on secularization has focused on whether and to what extent individual belief and practice have been declining [Norris and Inglehart Reference Norris and Inglehart2004; Stark and Finke Reference Stark and Finke2000], scholars have recently expressed renewed interest in the secularization of societal institutions. These authors have investigated the processes whereby specialized social institutions break free (“differentiate”) from religious authority, and whereby religious institutions come to develop and specialize in a specifically “religious” function [Casanova Reference Casanova1994; Reference Casanova2006; Gorski Reference Gorski2000; Reference Gorski, Adams, Clemens and Shola Orloff2005; Martin Reference Martin2005; Smith Reference Smith and Smith2003]. As part of this reevaluation, scholars have begun to explore how secularization can lead to different configurations and roles for religion in public life in different countries—what Gorski and Altinordu [Reference Gorski and Altinordu2008: 76] term “secular settlements.”
In his classic comparative analysis of secular settlements in the Christian West, Martin [Reference Martin1978] argues that secularization proceeded differently depending on whether a country was Catholic or Protestant, its degree of religious pluralism, and whether it had an established church. Several recent studies [Casanova Reference Casanova1994; Gorski Reference Gorski and Dillon2003; Reference Gorski, Adams, Clemens and Shola Orloff2005; Halikiopoulou Reference Halikiopoulou2011; Kuru Reference Kuru2009; Martin Reference Martin2005] have largely confirmed Martin’s argument, especially his finding that established churches tend to promote more secular outcomes by encouraging anticlericalism. Kuru [Reference Kuru2009], for instance, finds that the United States developed a “passive” form of secularism, in contrast to the staunchly anticlerical “assertive” secularism found in France and Turkey, because America lacked an Ancien Régime uniting church and state. Yet the United States and Australia do not differ meaningfully according to these criteria: both nations are historically Protestant yet religiously pluralistic, neither features an established church, and both share an attitude of “passive” secularism [Kuru Reference Kuru2009: 27]. The reason for their divergent secular settlements must be found elsewhere.
Alternative explanations can be found in two agent-centered approaches which focus on the interests and motivations of particular political actors. Gill’s [Reference Gill2008] rational-choice approach, for instance, argues that variations in “religious liberty” reflect the calculations of political actors motivated to preserve their political power, maximize economic resources, and ensure social stability. State policies toward religion are means to these ends, and change as necessary to grow and maintain political power. Divergent secular settlements should thus reflect these self-interested political calculations. By contrast, the secular movements approach argues that secularization is the result of deliberate campaigns by rising intellectual elites to enhance their social position at the expense of established religious elites [Smith Reference Smith and Smith2003]. By implication, variations in secular settlements should reflect the differential successes of these campaigns, as conditioned by access to resources and political opportunities.
While there is much to commend these agent-centered approaches, neither approach satisfactorily explains the American-Australian variation. Neither can convincingly explain why educators in the United States and Australia developed such consequentially different understandings of their professional interests, as we shall see that they did, or why they pursued such different reform strategies. In large part, this is because both approaches ultimately assume, or focus on, a very limited range of self-aggrandizing and anticlerical interests and motivations. As other studies have shown, however, the motives animating secularizing actors are quite variable [Kuru Reference Kuru2009; Mayrl Reference Mayrl2011], and the origins of those motives must be interrogated, rather than taken for granted. Further, the secular movements approach lacks comparative purchase. Because it was designed to explain the success or failure of particular campaigns in a single country, it is not particularly well-suited to explaining why or how political opportunities, resources, or interests might vary cross-nationally, or why secularizing actors should have arisen in some countries but not others.
One major reason why none of these approaches on its own offers a satisfying explanation is that they pay insufficient attention to the administrative and bureaucratic structure of the state. The state is a complex, internally differentiated amalgam of institutions that varies substantially from country to country [Evans, Rueschemeyer and Skocpol Reference Evans, Rueschemeyer and Skocpol1985]. These variable institutional configurations are an important potential influence that may contribute to the pace and extent of secularization. To the extent that existing approaches have examined the state, they have largely restricted their attention to formal religious establishments (state churches). As noted above, these studies find that established churches can generate anticlerical movements, which, in turn, promote more unyielding forms of secularity. Indeed, the presence or absence of state churches has been regularly invoked to explain the contrasting secularisms of Europe and the United States [Berger et al. Reference Berger, Davie and Fokas2008; Casanova Reference Casanova1994; Kuru Reference Kuru2009; Martin Reference Martin1978; Reference Martin2005]. Yet by focusing so narrowly on established churches, these studies overlook the way in which other aspects of the state may contribute to the rise of distinctive secular settlements. The established church, in other words, may be just one instance of a much wider array of structural variations within the state with important consequences for secularization.
Agent-centered approaches, for their part, have largely overlooked the state in their focus on individual motives. Although the secular movements approach notes that “state expansion” was a “facilitating structural force” abetting the secular revolution [Smith Reference Smith and Smith2003: 2], it leaves the particular ways that the state is linked to secularization implicit (but see [Thomas, Peck and De Haan Reference Thomas, Peck, De Haan and Smith2003]). Similarly, the rational-choice approach strongly implies that political institutions could constrain rational political actors, but these institutional constraints rarely feature in the empirical analyses [Gill Reference Gill2008]. Overall, the state appears only in fragmentary glimpses in these works; neither approach theorizes the actual administrative and bureaucratic structure of the state as a serious independent, variable factor that might contribute to diverse secular settlements.
Nevertheless, existing research does offer some tantalizing hints that state structure—and, in particular, the degree of administrative centralization—may relate to patterns of secularization. Martin [Reference Martin2005: 67], for instance, argues that the relative centralization of media and other institutions in Britain and Scandinavia has allowed “secular elites” to achieve greater “intellectual influence” in promoting their irreligious ideas. Yet although he called for further research into how administrative centralization relates to secularization, to date that call has not been answered. Some empirical evidence appears to corroborate the link between administrative centralization and secularization [Gill Reference Gill2008: 207-212], but other evidence suggests that secularization may be more closely linked to decentralized administration instead [Halikiopoulou Reference Halikiopoulou2011: 105-106; Monsma and Soper Reference Monsma and Soper1997: 177-180]. Because this evidence is mostly mentioned in passing, however, rather than explicitly theorized, there is little consensus on the relationship between the structure of state administration and secularization.
Secularization and the state in the United States and Australia
Australia and the United States are similar in several respects thought to influence secularization [Martin Reference Martin1978]. Both countries are predominantly Protestant with a sizable (∼25%) Catholic minority and substantial religious pluralism; neither country features a denomination with a majority market share [Australian Bureau of Statistics, 2006; Bader, Mencken, and Froese Reference Bader, Mencken and Froese2007]. Further, neither country has an established church; indeed, both nations have constitutions that formally disestablish religion in nearly identical terms.Footnote 1 Beyond these religious and constitutional commonalities, the countries are also similar in other important respects. By 1900, both nations were substantially modern both economically, with high gdps, significant industrial development, extensive rail networks, and predominantly urban populations; and politically, with federated, democratic political institutions [Archer Reference Archer2007]. Each country also enrolled more than two-thirds of its children in primary education from the 1870s onward [Benavot and Riddle Reference Benavot and Riddle1988]. Finally, religion was a traditional part of education in both nations. In the mid-to-late nineteenth century, each nation permitted religious instruction in its public schools (although to varying extents in different states), while broadly prohibiting financial support for religious schools [Mayrl Reference Mayrlforthcoming].
Religious education in the United States and Australia
Despite these similarities, religion’s role in public education evolved in different directions in the late nineteenth and early twentieth centuries. Even before the Supreme Court declared public school devotionals unconstitutional in the early 1960s,Footnote 2 the “nonsectarian” devotional exercises that had been the mainstay of the nineteenth-century American common school were in decline. This decline is dramatically visible in Bible reading, traditionally the most important component of American religious education [Moore Reference Moore2000]. Official surveys of school districts in 1896 and 1903 found that approximately three-quarters of school districts featured Bible reading [United States Office of Education 1904]. Some 60 years later, however, a similar survey found that Bible reading took place in just 42% of schools [Dierenfield Reference Dierenfield1962]. This decline, depicted graphically in Figure 1, occurred despite a concerted effort to pass mandatory Bible reading laws in many states, as discussed below [Tyack, James and Benavot Reference Tyack, James and Benavot1987].
Figure 1 Bible reading in American public schools, 1896-1987
Note: The vertical dashed line indicates the date (1963) of the Abington v. Schempp decision of the United States Supreme Court prohibiting Bible reading in public schools.
Sources: United States Office of Education [1904: 2445)], Dierenfield [Reference Dierenfield1962: 51; 1967: 447-448; 1986: 391], Way [Reference Way1968: 199].
In Australia, by contrast, religion’s position in the schools remained secure well into the 1960s. Australia featured a dual system of religious education: teachers in most states were required by law to provide “general religious teaching,” including readings from official, state-sanctioned Scripture readers; and clergy were also given “right of entry” to visit the state schools once a week in order to provide supplementary “special religious instruction” (sri) to the children of their denomination [Langdon Reference Langdon1986]. While exactly comparable survey data on practices do not exist for Australia, a comparison of legal provisions in 1880 and 1950 reflects the persistence and even entrenchment of this dual system.Footnote 3 As Table 1 demonstrates, the state-level variation regarding religious exercises in the late nineteenth century converged over the first half of the twentieth century. By 1950, sri was essentially a universal practice throughout the country, and most Australian states also mandated Scripture reading.
Table 1 Religious education provisions in the Australian states, 1880-1950
Sources: Barcan [Reference Barcan1965: 134-137; Reference Barcan1980:156-157], Wigney [Reference Wigney1958].
In short, religious education declined slowly in American schools over the late nineteenth and early twentieth centuries, whereas it retained a strong position in Australian education well into the 1960s. In the remainder of this paper, I will argue that these divergent trajectories can be accounted for in large part by still another difference between Australia and the United States—namely, the structure of each nation’s system of educational administration.
The administrative context of secularization: permeable and insulated states
Australia and the United States feature considerably different systems of educational administration. In the United States, education was highly decentralized [Reese Reference Reese2005]. Administrative control over schools—including hiring and firing of personnel, curricular decision-making, and supervision of teachers—was delegated to local school districts [Cremin Reference Cremin1980]. In larger communities, school districts often hired a superintendent of schools to oversee and coordinate multiple schools within a district, but these superintendents were fundamentally local officials [Justice Reference Justice2005]. State departments of education, by contrast, were comparably quite weak, with some powers of oversight and coordination but little to no control over personnel or curricula [Tyack et al. Reference Tyack, James and Benavot1987]. Accordingly, educational policy and practice was riotously diverse, varying not only from state to state, but from district to district within each state.
Australian education, by contrast, was quite centralized. All education was coordinated through central bureaucracies operating out of each state capital [New South Wales, Department of Education 1978]. Education was run as part of the general civil service, meaning that both teachers and administrators were bound by public service regulations [Pike Reference Pike1965]. Local schools, and the headmasters who managed them, were responsible not to local communities, but to departmental officials, who were in turn ultimately responsible to elected officials in the state legislature. Curricula, policies, and teaching techniques were prescribed and monitored from the center through an efficient and exacting cadre of inspectors [Cleverley and Lawry Reference Cleverley and Lawry1972]. Educational practice was far more uniform as a result.
Table 2 Key institutional features of the state relevant to secularization
These two administrative systems, illustrated schematically in Table 2, had important consequences for secularization. America’s decentralized system, with its widely dispersed authority and policy-setting structures, amounted to a permeable state, featuring many access points where political challenges could take place and new ideas could take root. Australia’s centralized system, by contrast, with its highly centralized authority and policy-setting structures, amounted to an insulated state, where any changes or challenges required the persuasion or cooptation of the entire state government. These two state structures shaped the ability of would-be reformers to challenge religious instruction in the late nineteenth and early twentieth centuries.Footnote 4
The state as institutional mediator of secularization
The first way that the state structures secularization is by acting as an institutional mediator of political conflicts between groups with divergent opinions about the appropriate relationship between religion and society. Political institutions typically grant greater access, leverage, and decision-making power to some groups rather than others, and the structure of the state regulates the ease with which political actors can gain access to decision-makers [Skocpol Reference Skocpol1992; Weir and Skocpol Reference Weir, Skocpol, Evans, Rueschemeyer and Skocpol1985]. Decentralized states typically contain more of these access points where political decisions may be contested. In democratic polities, moreover, because local officials are responsible and accountable to more locally-defined populations, they have incentives to heed calls for policy change by their constituents. Decentralized states are therefore more readily influenced by groups seeking to alter religious education policy, particularly in situations where challengers are geographically concentrated. Further, decentralization can allow challengers to triumph in some locales even if they do not triumph everywhere. Accordingly, several studies have shown that religious claimants’ influence on policy is enhanced in relatively decentralized states [Fetzer and Soper Reference Fetzer and Soper2005; Gill Reference Gill2008].
Generally speaking, conflicts over religious policy may be waged among religious, antireligious, and/or irreligious groups [Gorski Reference Gorski and Dillon2003; Mayrl Reference Mayrl2011]. In Australia and the United States, however, the most important conflict mediated by the state in the late nineteenth and early twentieth centuries was between Catholics and Protestants.Footnote 5 In both nations, an informal Protestant establishment dominated public life [Handy Reference Handy1991; Hogan Reference Hogan1987]. Catholics and other religious minorities challenged these establishments, especially in public education. While the “nonsectarian” religious instruction in the public schools satisfied Protestants, Catholics found its form and content to be objectionably and insidiously Protestant. In both countries, therefore, Catholics took a hard line against religious exercises in the public school, and—joined at various times and in various places by other religious outsiders—spearheaded campaigns challenging the “nonsectarian” religious education policies that the Protestant establishment had devised [Austin Reference Austin1965 [1961]; Hennesey Reference Hennesey1981].
The degree of centralization of educational administration was pivotal in determining the outcomes of these struggles. Because administrative control was centered in local school boards in the United States, those boards became battlegrounds for decisions about religious education. Decentralization thus set in motion two primary secularizing mechanisms. First, because Catholics and other religious minorities in the United States clustered in larger cities [Finke and Stark Reference Finke and Stark1988], they could amass sufficient political influence to capture or co-opt local boards, leading to policy change. Beyond this dynamic of institutional capture, however, the larger number of institutional access points meant that conflict over religious education became a regular and highly-visible feature of American political life. In response, boards and administrators developed a new administrative attitude toward religious instruction that viewed it as problematic and divisive. In Australia, by contrast, the centralization of administrative authority at the state level precluded the development of these mechanisms.
Local control and institutional capture in the United States
America’s decentralized system gave religious minorities ample opportunities to challenge, and sometimes overturn, the “nonsectarian” pan-Protestant devotionals in the public schools. As communities grew and diversified religiously, local school boards diversified as well, making them increasingly open to the arguments of those who claimed that Bible reading was an essentially Protestant activity with no place in the supposedly common public schools. Perhaps the most consequential incident of this type took place in Cincinnati between 1869 and 1873 [Green Reference Green2010]. In 1869, the city’s Board of Education voted to prohibit Bible reading in the city schools, with all Catholic board members voting in favor, and the majority of Protestant members voting against. Their decision provoked a national uproar. Furious Protestants took the Board to court, but the Ohio Supreme Court upheld the Board, ruling that its decision was an appropriate exercise of local authority.Footnote 6
The “Cincinnati Bible War” instigated pro- and anti-Bible reading campaigns that turned Bible reading into a major issue for school boards nationwide in the following decades. While many challenges were resolved in favor of Bible reading, religious minorities were able to notch victories in several religiously diverse cities, including Buffalo, Rochester, and Chicago, where they were able to persuade or pressure local school boards to change their policies [Green Reference Green2010; Justice Reference Justice2005]. Although Catholics were at the leading edge of the charge, Jews also became increasingly outspoken in their opposition to Bible reading after 1900. In 1905, Jewish groups launched a broad public campaign against religious devotionals with a widely distributed tract entitled “Why the Bible Should Not Be Read in the Public Schools,” and successfully argued against Bible reading in the Washington, DC schools [Cohen Reference Cohen1992]. By the outbreak of World War I, these local protests were increasingly displacing Bible reading even in relatively smaller cities [Adler Reference Adler1917].
Local control, however, was a two-way street, and the dynamics of institutional capture could work for Protestants as well as Catholics. During the nineteenth century, Protestants leveraged local control to maintain religious devotionals in several jurisdictions [Green Reference Green2010], and in the 1920s nativist groups successfully pressured local school boards in the South and Midwest to reintroduce Bible reading in some locales where it had previously been eliminated [Laats Reference Laats2010]. As local challenges to Bible reading mounted, however, defenders increasingly took steps to take the decision out of local hands so it would be less susceptible to political pressure from restive opponents. In particular, Protestants in the early twentieth century attempted to entrench devotionals by appealing to state officials to impose mandatory Bible-reading legislation. Between 1913 and 1930, eleven states and the District of Columbia passed such laws [Tyack et al. Reference Tyack, James and Benavot1987: 165]. By enshrining the practice in state law, these campaigns took the issue away from local authorities, effectively foreclosing institutional capture (of local boards) as a means of policy change.
From controversy to conventional wisdom: religion as administrative problem
While local control directly facilitated secularization by permitting institutional capture, it also contributed more indirectly by generating a climate of administrative wariness toward religion. Decades of ongoing local conflict over religion created an understanding among principals, superintendents, and school board members that religion was an inherently controversial issue, and therefore a potential administrative headache. By the late nineteenth century, the idea that religion would be likely to do more harm than good, from an administrative perspective, began to consolidate into conventional wisdom. This wariness of controversy, even in places where political dynamics might have favored Protestant successes, often convinced administrators, concerned above all with minimizing conflict and ensuring the smooth operation of their school systems, to exclude religious exercises.
Local school boards often explicitly evoked the threat of religious controversy to justify their actions, as in New Haven, where the Board of Education discontinued religious exercises in 1877 after declaring that the prospect of sectarian conflict in their school system required “its entire secularization” (quoted in [Mason 1953: 162]). As objections mounted, the Bible no longer seemed unifying and “nonsectarian”—instead, it seemed vaguely menacing. By the turn of the century, repeated episodes of local conflict over religious instruction had seeped into professional discourses as a cautionary form of administrative wisdom. “It is impossible to have any such unsectarian religion that is not regarded as sectarian by the more earnest religious denominations,” declared US Commissioner of Education William T. Harris in 1903, in discouraging the use of devotional practices [Harris Reference Harris1903: 226-227]. A few years later, a member of the New York City board of examiners noted that dropping Bible reading to avoid controversy was a widespread practice: “If any form of exercise is found to give offense to any, our procedure is rather to cut out that exercise, not to try to adapt it to differentiated groups” [Hervey Reference Hervey1907: 82]. Localized religious conflict thus created an amplified, self-propelling echo in the form of arguments against religious education grounded in fear of religious controversy. By the 1910s, many administrators and educators were inclined to view religious instruction through the lens of controversy, which in turn predisposed them to downplay or eliminate Bible reading altogether [e.g. Wild Reference Wild1916].
In short, local control of educational policy in the United States abetted secularization. By creating thousands of autonomous sites where religious education policy was decided and contested, decentralized education enabled critics to challenge local board policies, contributing directly to a decline in religious education; and generated widespread conflict over religious education, indirectly encouraging educational administrators to drop Bible reading out of fear of controversy. That local control contributed in these two ways to the slow erosion of religion’s place in the American public school can be seen when contrasted with Australia, where these conditions did not hold.
Centralized control and the suppression of religious conflict in Australia
In Australia, these kinds of local challenges to religious education were simply not possible. Under the centralized Australian system, there were no local boards to petition for redress. Nor were local administrative officials, the school headmasters, given any policymaking authority. Accordingly, all local requests for alterations to official department policy had to be run through the head office. Educational officials had few incentives to take external petitions seriously, and tended to dismiss them as a matter of course. When Catholics complained about offending religious practices, therefore, they were typically rebuffed. In New South Wales, for instance, the Catholic Federation lodged a complaint in December 1920 against the use of the Protestant Lord’s Prayer to open school days, and asked that the Catholic version be made available as an alternative. The department replied that it viewed the difference between the two versions as being of “little consequence,” and refused to alter its policy [Board Reference Board1921]. Similarly, when Catholics petitioned to remove several anti-Catholic statements from history textbooks, department officials denied nearly every claim, dismissing their complaints as “ridiculous” and “trivial” [Cramp and Smairl Reference Cramp and Smairl1921].
Centralization thus deprived Catholics and other religious minorities of much of the leverage that local control provided their American counterparts. Because department officials were ultimately beholden to the legislature, their actions were shaped by the values and wishes of the broader community, which was overwhelmingly Protestant, rather than by any local dynamics which might have been more favorable to minorities. Department officials were typically quite cozy with Protestant leaders, often expressing their support for Protestant aims and even colluding with Protestant leaders to expand religious education. Officials in Sydney informally encouraged schools to perform worship services at the start of the school day, even though the practice was not formally sanctioned by law [Council for Christian Education in Schools c.1952]; and proclaimed their “warm sympathy” for a group of Protestant clergymen’s efforts to improve religious instruction in the schools [Pelham Reference Pelham1952].
In such a context, Catholics and other religious minorities, lacking any local recourse, could do little but rail against the state government. In 1935, the Catholic Bishop of Goulburn told members of the Holyname Society: “An appeal to reason, to justice, and fair play is useless where prejudice prevails, and it is therefore necessary for you, the Catholic men of N.S.W., to appeal in the only way that has any telling effect—that is, the appeal of the ballot box” [Labour Daily 1935]. Unfortunately for Catholics, their inadequate numbers meant that even that appeal resulted in little redress before the 1960s. Centralized control effectively eliminated their leverage to contest the Protestant devotionals that characterized Australian schools well into the twentieth century.
The state as constitutor of secularizing interests
A second way that the state structures secularization is by creating structural conditions that foster the emergence of pro-secularizing interests. The relations of authority embedded in political institutions encourage actors to construe their interests in particular ways [Halfmann Reference Halfmann2011]; while administrative rules and regulations shape both how actors organize and what kinds of ideas they articulate, develop, and disseminate. In this regard, states shape the interests, preferences, and beliefs of the actors who inhabit them, ultimately helping to generate the very actors who subsequently seek to secularize them.
In the field of education, the most important “secularizing” group promoted by the state was professional educators. The development of new, professional jurisdictions capable of making authoritative knowledge-claims has been identified as a particularly important secularizing force [Chaves Reference Chaves1994; Dobbelaere Reference Dobbelaere1981]. In education, the late nineteenth century saw the rise of a new “science of education” which supplanted the traditional, religiously-based forms of knowledge with novel, non-religious ones [Beyerlein Reference Beyerlein and Smith2003; Smith Reference Smith and Smith2003]. As education developed into a distinct profession, it also developed a new pedagogical orientation geared toward science, critical thinking, and the process of learning [McClellan Reference McClellan1999]. These ideas and practices increasingly crowded religion out of an expanding curriculum and called religion’s educational value into question. In the process, religion became almost entirely marginal to professional understandings of education.
The secular movements approach to secularization has provided great insights into how these new professional ideas transformed religion’s role in American education [Beyerlein Reference Beyerlein and Smith2003; Smith Reference Smith and Smith2003; Thomas et al. Reference Thomas, Peck, De Haan and Smith2003]. In particular, they have shown that professionalization was advanced by educational elites for whom the new science of education promised increased authority, legitimacy, and job security. Inspired by the Enlightenment’s faith in reason, science, and the power of the state to mold democratic citizens, these elites used professional organizations as mobilizing vehicles from which to promote their ideas and delegitimize existing practices. Professionalization, in short, was also a political project whose success hinged on sidelining existing educational elites, and the religious beliefs and practices that sustained their legitimacy and authority.
What the secular movements approach largely elides, however, is the extent to which the decentralized American state, and the professional infrastructure that developed in response to it, made this movement possible (but see [Thomas et al. Reference Thomas, Peck, De Haan and Smith2003]). Many of the most prominent professionalizing educators were superintendents and other state actors. The incentives and opportunities generated by their position within the state were essential both in shaping their interests, and in creating spaces in which their ideas could develop and their campaigns could mobilize. The importance of this oversight becomes clear in comparative perspective. The professionalization of educators did not take place to the same extent in Australia and the United States, and consequently educators were not a major secularizing force on Australian education [Cleverley and Lawry Reference Cleverley and Lawry1972]. Whereas American teachers adopted a professional outlook with beliefs and practices that directly contributed to secularization, Australian teachers adopted an industrial outlook which did not challenge religion’s role in education.
In this section and the next, I argue that the structure of state administration helped constitute different interests and actions among educators in the two countries. The decentralized American state created incentives that fostered the emergence of a professional infrastructure in which professional ideas and identities could be developed, and generated structural positions from which secularizing ideas and practices could be implemented. In Australia, by contrast, the centralized state actively discouraged the development of this infrastructure; pushed Australian teachers toward an industrial, rather than professional, identity; and prevented the spread of innovative pedagogical reforms.
Decentralization, professional infrastructure, and secularization in America
American education professionalized early thanks to its vibrant professional infrastructure—that is, its array of professional journals, associations, and other mechanisms of knowledge exchange. Educational periodicals, which first appeared in the 1810s, multiplied dramatically in the last quarter of the nineteenth century. Many of these journals dedicated themselves to encouraging sustained debate over educational issues, and many of the first ideas challenging received pedagogical orthodoxy were introduced to American educators through these journals [Davis Reference Davis1970 [1919]]. Perhaps more importantly, American educators also developed strong educational associations at the local, state, and federal levels [Fraser Reference Fraser2007]. These associations were important forums where debates over new educational ideas, practices, and approaches could take place; and they also played a crucial role in disseminating these ideas through free pamphlets and other publications [Mattingly Reference Mattingly1975].
This elaborate professional infrastructure developed in large part in response to the decentralized nature of American education. Because state educational agencies were weak, the creation of common schools was typically left to local initiative [Kaestle Reference Kaestle1983]. American educators thus formed professional associations and founded educational journals as a means of promoting public education [Mattingly Reference Mattingly1975; Tyack and Hansot Reference Tyack and Hansot1982]. State superintendents, meanwhile, with few resources and highly circumscribed formal powers, encouraged the growth of these professional institutions as a means both of promoting education, and of extending their own influence. For instance, in an effort to increase the power of his office, California’s Superintendent of Education pushed for enhanced training and certification of teachers, and distributed a state-funded educational journal to teachers and school trustees around the state [Tyack et al. Reference Tyack, James and Benavot1987]. In many states, the first steps to create educational journals were taken by state superintendents or commissioners of schools, who saw these journals as an important means of communicating their own opinions to a broader audience [Davis Reference Davis1970 [1919]].
Indeed, something of a symbiotic relationship developed between teachers’ associations and state superintendents, as each used the other to advance their goals and interests. A particularly intimate example of this took place in Illinois. There, the Secretary of State, who also served as ex officio Superintendent of Common Schools, met with several school principals to form the Illinois State Teachers’ Association (ista) so that they might better agitate on behalf of public education. Two years later, the ista successfully lobbied for the creation of an independent Office of State Superintendent [Bone Reference Bone1957]. Throughout the mid-nineteenth century, these symbiotic ties were evident in state educational associations, which “tended to be dominated by the state superintendents, who used them to extend their own influence in standardizing educational practice” [Tyack and Hansot Reference Tyack and Hansot1982: 49].
With local boosters building a profession from below, and state officials nurturing it from above, a vibrant educational infrastructure proliferated in the United States. The result was an increasingly interconnected network of journals and associations, largely independent of local educational authorities, in which novel educational ideas and practices could be raised, debated, and disseminated; and where a professional identity and set of professional standards could develop. Through this infrastructure, new scientific and progressive approaches to the theory and practice of education insinuated themselves deeply into the educational community. And as these new ideas were elaborated, traditional religious education was often marginalized or displaced.
One important site where professionalization sped religious decline was the National Education Association (nea). As Beyerlein [Reference Beyerlein and Smith2003] has demonstrated, the nea, with its national scope and large membership, was a crucial forum in which nineteenth-century educators debated the appropriate role for religion in the public schools. While nea leaders initially favored Bible reading in the mid-nineteenth century, religion’s place in the public school curriculum became a central topic of debate within the organization. By the turn of the twentieth century, the argument that religion was unscientific and unsuited for modern society gained favor, leading the nea to pass a resolution in 1902 calling for the Bible to be read as literature rather than as a “theological book” [Report of the Committee on Resolutions 1902: 27]. Further debate on the issue ultimately led the nea to reject the very idea of using revealed religion in the classroom. By 1937, official nea policy called for “the utmost possible emancipation from the dictates of a priori or dogmatic notions, whether of theological revelation, Colbertian mercantilism, Ricardian individualism, or Marxian communism” [Educational Policies Commission 1946: 48].
Nor was the nea the only professional site where religion was marginalized. Ironically, the embrace of novel educational ideas is perhaps most dramatically visible in the Religious Education Association (rea), a forum founded in 1903 for the explicit promotion of religious education. Articles in the rea’s journal, Religious Education, meticulously tracked how new developments in psychology and pedagogy affected religious education, with most concluding that these developments required the wholesale transformation of religious education and the abandonment of devotional Bible reading. Above all, traditional religion would have to be subordinated to science. The US Commissioner of Education advised rea members in 1907 that since “modern education is allied with modern science […] we may confidently expect that in this age it will mold religious education to its standards and processes” [Brown Reference Brown1907: 121]. This molding typically involved stripping religious education of any supernaturalistic or revealed qualities. In 1919, the rea’s president admonished readers that education should not “encourage the delusive belief in supernatural agencies and dependence upon them, but it should be such as to convince everybody that things can be controlled and moulded by the power of man” [McGiffert Reference McGiffert1919: 157]. By the early twentieth century, therefore, professional associations like the nea and rea were nurturing and promulgating a professional consensus in favor of a less traditional, more scientific, and secular curriculum that left little room for devotional Bible reading.
Centralization and stasis in Australia’s anemic professional infrastructure
In Australia, by contrast, a comparably robust network of associations, journals, and other professional institutions never developed. In stark contrast to the burgeoning American periodical scene, only a handful of periodicals were operating in Australia at the turn of the twentieth century, most of which were official organs of the state education departments and (for reasons elaborated below) not sites for autonomous discussion of educational ideas. Moreover, while Australian teachers began to organize into unions in the late nineteenth century, their associational life apart from the unions was anemic. When American educationist Isaac Kandel [1938: 81] visited in 1937, he expressed dismay at the “almost complete absence of associations and societies for the study of education […] paralleled by an absence of professional journals except those published by the teachers’ associations and unions.”
The absence of a robust professional infrastructure in Australia was not mere accident, but rather reflected constraints placed upon teachers by the centralized educational system. Centralized controls both encouraged Australian teachers to adhere to existing policies and practices and discouraged them from critically discussing them. Educational orthodoxy was regularly reinforced by a cadre of official inspectors, who visited each school to examine students and observe teachers. The inspectorate was a staunchly conservative force within Australian education, focused on ensuring that teachers adhered closely to official syllabi and departmental regulations [Cleverley and Lawry Reference Cleverley and Lawry1972]. Visitors regularly noted that the inspectorate looked with disapproval upon teachers displaying pedagogical initiative, thereby stifling innovation and experimentation. Kandel [Reference Kandel1938: 62] observed that “Originality and initiative are discouraged, and a teacher or headmaster who introduces some experiment or innovation may even be written off by an inspector for ‘showmanship’.” This enforced conformity extended to religious instruction as it did to other parts of the curriculum; well into the 1960s, district inspectors ensured that Scripture was being taught according to the approved syllabus—exactingly, in some cases [Burns Reference Burns1963: 260-261].
Further, as civil servants, Australian teachers were subject to restrictive public service regulations which shackled their ability to speak out on educational matters. Teachers were barred by law and regulation from publicly commenting upon departmental administration, and these regulations were wielded by department officials on an irregular but consistent basis to keep teachers in line. In the 1870s, for instance, teachers in New South Wales started a teacher’s journal with the consent of department officials. But when they published an article calling for the legislature to pay closer attention to teachers’ concerns, department officials shut down the journal, suspended the authors from their teaching duties, and exiled the journal’s editor to a remote Outback school as punishment [Pike Reference Pike1965]. Even though enforcement of such regulations waned over time, the threat of their enforcement had a chilling effect both on teachers’ willingness to criticize the system, and on their ability to organize. As Mitchell [Reference Mitchell1975: 89] notes, the regulations were “an excuse for the timid to remain silent. Most teachers had neither the desire nor the courage to voice their grievances publicly.” The associations and journals that did develop, therefore, in addition to being less numerous, were also almost wholly uncritical of existing educational practices. Instead, they focused on “matters affecting the status of teachers, fresh regulations, news, and occasionally notes of lessons on the courses of study as they are” [Kandel Reference Kandel1938: 81].
These rigid controls not only discouraged Australian teachers from discussing educational problems or experimenting with potential reforms, it also affected how they organized. Unlike their American counterparts, who organized on a professional basis, Australian teachers organized instead on an industrial basis, as unions. Although their rights regarding criticism of department policies may have been curtailed, their rights under trade union laws provided a sounder legal basis for organizing [Spaull Reference Spaull1986]. Yet organizing on this basis meant that Australian teachers’ groups focused heavily on wages and working conditions, rather than on pedagogy or curricular reform. The constitution of the New South Wales Teachers’ Federation (nswtf), for instance, discouraged consideration of curricular issues; conference discussions throughout the 1920s focused on arbitration, service conditions, and salaries to the near-total exclusion of educational matters [Mitchell Reference Mitchell1975].
This overwhelmingly industrial focus drew frequent criticism. Visiting Australia in 1955, American educational historian R. Freeman Butts [1955: 79] lamented that “Australian teachers, imbued with trade union tradition, are under-organized with respect to professional stimulation, exchange of ideas, and mutual criticism.” As late as 1975, an observer noted that “teachers’ unions have been criticized for directing their activities mainly towards improving salaries and conditions of service […] without making any notable corresponding efforts towards professionalizing the service, particularly by helping to improve educational practices and to upgrade professional expertise among teachers” [Maclaine Reference Maclaine1975: 127]. Educators’ industrial focus thus functioned throughout the first half of the twentieth century to render the curriculum essentially free from criticism or experimentation by professional educators. Not surprisingly, the curriculum changed very little until the 1960s [Connell Reference Connell1993].
This broad stasis in curriculum and pedagogy was as evident in religious education as it was in other areas. When a radical communist faction developed in the nswtf in the late 1920s, it initially called for the abolition of religious education. However, as its leaders gained power within the Federation in the 1940s, it dropped these radical demands in the face of widespread disagreement on the part of teachers, who evinced “indifference or conservatism about the quality of education, and [instead] anxiety about salaries” [Mitchell Reference Mitchell1975: 173-174]. Indeed, a 1963 analysis of nswtf policy toward religion concluded that, largely because “very little discussion” of religious education had taken place, “nothing has been added to thought on education in religion and morals by the Teachers’ Federation” [Burns Reference Burns1963: 136-137]. When it came to secularizing professional reforms, therefore, Australian teachers were simply not major advocates in the years before 1960.
The state as focusing mechanism
Certain administrative offices may be especially conducive to secularization because they combine mediating and constitutive elements in the same structural location, both generating secularizing agents and allowing them to disseminate reforms in a particularly effective manner. Where they exist, these dynamic institutional loci act as focusing mechanisms that magnify the effect of secularizing processes. Importantly, in the United States such a focusing mechanism—the city superintendency—existed, whereas no parallel office existed in Australia. This allowed inherent secularizing trends to proceed still faster in the United States than in Australia.
City superintendents, who coordinated the management of urban school districts, had incentives to promote the development of professional standards and knowledge about education [Thomas et al. Reference Thomas, Peck, De Haan and Smith2003]. By arguing that “only experts […] could keep apprised of the latest pedagogical ideas through wide reading, professional correspondence, and association with other urban leaders,” city superintendents could press for greater power and autonomy from the local boards of education who employed them [Reese Reference Reese2005: 60]. Superintendents were thus at the forefront of the movement for the development of new professional expertise to support these claims. Because of their personal interest in advancing professional development, superintendents often took key leadership roles in professional associations such as the nea [Beyerlein Reference Beyerlein and Smith2003; Tyack and Hansot Reference Tyack and Hansot1982]. They also served as editors of prestigious educational journals, and as important advisors to the United States Commissioner of Education, in the late nineteenth and early twentieth centuries [Callahan Reference Callahan1964].
More importantly, because superintendents were situated at a unique crossroads—as administrative leaders at the local level, and as members of an emergent educational policymaking elite at the national level—they were strategically positioned to simultaneously develop and enact professional reforms. Enterprising urban superintendents introduced a range of scientific and progressive reforms on a piecemeal basis in the late nineteenth and early twentieth centuries [Cuban Reference Cuban1993]. Because superintendents were administrators and not elected officials, they were somewhat insulated from political pressures, allowing them to function as conduits for secularizing professional norms, attitudes, and reforms [Thomas et al. Reference Thomas, Peck, De Haan and Smith2003]. And local control meant that such reforms could spread even if entire states did not adopt them as official policy.Footnote 7
Educational reforms often displaced the traditional materials, methods, and organization in which religious exercises were embedded, without providing the same—or any—role for religion in the new, reformed system. For example, when Robert and Helen Lynd visited Muncie, Indiana, in the mid-1920s to research their classic study Middletown, they observed a public school system rich with religious elements and a curricular philosophy that declared that “pupils should learn to appreciate the Bible as a fountain of truth and beauty” (quoted in [Lynd and Lynd Reference Lynd and Lynd1929: 204]). When they returned a decade later, however, the picture had changed dramatically. Their description of the public schools is nearly bereft of any reference to religion, and is instead given over to a discussion of the impact of a “ten-year program of school planning and reorganization” that had been implemented in 1928. This reorganization entailed nothing less than “the redefinition of the philosophy of education in Middletown” with a new emphasis on “individual differences as over mass education and conformity.” The Lynds illustrated this transformation by quoting from a 1933 district planning report that rejected the idea that “to learn is basically acquisition and acceptance on authority” in favor of an educational philosophy of “equal opportunity for every child to develop according to his abilities, interests, and aptitudes” [Lynd and Lynd Reference Lynd and Lynd1937: 219-221]. As in Muncie, so in other cities across the United States: professional reforms were implemented piecemeal locally through the actions of reforming superintendents and school administrators.
In Australia, by contrast, there was no analogue to the American superintendency. Centralized administrative control meant that local administration was entrusted, not to quasi-independent superintendents, but instead to headmasters. Department officials selected as headmasters only those teachers who expressed enthusiastic support for the status quo, leading to “a promotion system which […] reaped its harvest in martinets” [Mitchell Reference Mitchell1975: 22]. And, since headmaster was often a springboard for still higher administrative office, headmasters had career incentives to toe the official line rather than engage in any experimentation. Accordingly, headmasters were a staunchly conservative lot, standing with administrators on behalf of departmental traditions and using their positions to discourage experimentation by teachers [McLean Reference McLean1955]. They were also highly supportive of religious education; in New South Wales, the Headmasters’ Association passed resolutions supportive of religious instruction well into the 1950s [Education 1956; New South Wales Teachers’ Federation 1959]. Far from acting as a focusing mechanism, then, the headmaster position served as a bulwark for traditional practices. Not surprisingly, this ensured remarkable continuity and uniformity in educational practice.
Discussion
While it would not be accurate to say that the state causes secularization, its administrative structure can decidedly shape the speed at and force with which secularization occurs. In the United States, decentralized educational administration facilitated challenges to religious exercises by religious minorities, fostered professional development and leadership among educators, and allowed novel educational practices oriented in new educational theories rather than religion to spread. In Australia, centralized state control over education insulated majoritarian religious exercises from criticism, suppressed professional development, and helped maintain the traditional educational practices that sustained religious instruction. These different dynamics help explain why religious exercises declined in American schools but not in Australian ones.
The state thus structures secularization in both mediating and constitutive ways. On the one hand, political institutions vary in their accessibility and responsiveness, generating distinct dynamics of institutional capture and administrative culture that mediate political conflicts over religious policy. On the other hand, the dynamics of authority and rules for behavior embedded in political institutions provide incentive structures that can help constitute the interests, ideas, and associational behavior of actors, including some who subsequently seek to secularize it.
This finding has several important implications for secularization theory. First, it opens new avenues for research into how the interests and motivations of secularizing actors may have been constituted, in whole or in part, by the state. Contemporary agent-centered theories of secularization have had relatively little to say about how the state shapes actors’ interests. The rational-choice approach [Gill Reference Gill2008] assumes that actors share a fixed set of interests, while the “secular movements” approach [Smith Reference Smith and Smith2003]—although acknowledging the complexity of secularizing actors’ motivations—has generally not interrogated these motives in light of their relationship to the state. This study suggests that this oversight may be consequential. That educational professionals were an important force in the secularization of American education has been recognized [Beyerlein Reference Beyerlein and Smith2003], but the existence of a professional identity and secularizing interests among educators is not inevitable, as the Australian experience demonstrates. In comparative perspective, secularizing campaigns are always embedded in particular political institutions, and we must be sensitive to how diverse cross-national outcomes might be accounted for by essentially institutional explanations [Kriesi Reference Kriesi, Snow, Soule and Kriesi2004]. Attending to the state sheds important light on dynamics of power and mobilization, and will therefore be a key step in any comparative extension of the secular movements approach.
Second, this study suggests a different interpretation of the relationship between administrative centralization and secularization. To the extent that this question has been posed, scholars have suggested that greater centralization is more conducive to secularization [Martin Reference Martin2005]. Certain familiar instances (e.g. France) come to mind which seem to fit this pattern. Yet by examining this relationship more directly, this study suggests that this is far from a universal relationship: the relatively centralized Australian educational system was more resistant to secularization than was the highly decentralized American one. The relationship between state centralization and secularization must therefore be qualified and contextualized.
One possible explanation for the unexpected relationship found in this study is that secularization from above has a different relationship to state centralization than secularization from below. Studies suggesting a positive relationship between state centralization and secularization typically view secularization as an elite project, in which elites can capture or use centralized institutions as a means of disseminating their secular ideas and policies [Berger et al. Reference Berger, Davie and Fokas2008; Martin Reference Martin2005]. Yet secularization is not only an elite project; it can also result from grassroots struggles by competing groups operating in localized contexts. Secularization, in other words, can also be advanced from the bottom up, for example, through religious conflicts.
Indeed, the fact that secularization also occurs from below helps account for the findings of this study. Precisely because the structure of the state is determinative in issues of democratic accountability, policy implementation, and the resolution of conflicts, decentralized administration can create opportunities for secularization from below. Thus, America’s decentralized school boards were a key institutional context through which religious pluralism led to more secular outcomes—and their absence in Australia helps explain the greater difficulties would-be bottom-up secularizers faced in that setting. By contrast, elite secularizing projects can be advanced through the “capture-the-flag” dynamics inherent in administrative centralization. Gaining control over centralized parties or departments can allow for rapid implementation of secular policies, as occurred in France [Kuru Reference Kuru2009]; and it can also allow for the gradual spread of irreligious ideas and practices, as has been noted in recent years in Britain [Berger et al. Reference Berger, Davie and Fokas2008]. To be sure, the particular mechanisms linking secularization from above and below with particular secular settlements require further clarification. Understanding under what circumstances centralization does and does not facilitate secularization is an important, yet still quite open, question for future research.
Finally, this study has implications for broader debates within the sociology of religion on the relationship between religious diversity and religious vitality. One of the core arguments of the supply-side theories of religion has been that religious diversity, in contexts of limited religious regulation, provides a spur to individual religiosity through mechanisms of religious competition [Gill Reference Gill2008; Stark and Finke Reference Stark and Finke2000]. This study suggests that, paradoxically, religious diversity can also drive macrostructural secularization even as it simultaneously supports ongoing religious vitality. This is something of a lost insight, despite the importance many studies of secularization qua differentiation have placed on how religious pluralism, and the religious conflicts that flow from it, can contribute to more secular outcomes [Gorski Reference Gorski and Dillon2003; Martin Reference Martin1978]. Given the complex, multidimensional nature of secularization [Chaves Reference Chaves1994; Dobbelaere Reference Dobbelaere1981], it is important to bear in mind that religious pluralism can have contradictory effects on secularization along its different dimensions.
The United States, here, is a case in point. Viewed from this perspective, the paradox of America’s remarkable institutional secularity ceases to be such a paradox. Instead, it reflects the divergent effects of religious pluralism and competition—as refracted through the decentralized administrative systems which structured those conflicts. Religious diversity and constitutional restrictions were important, but not determinative. The role of the state, instead, was key—both in structuring the conflicts over religious education that religious diversity helped bring about, and also in actively helping to generate sectarian and professional conflicts. America’s contemporary secular settlement in education thus reflects the democratic, permeable character of American educational administration as much as these other factors.
Conclusion
This article has demonstrated how and why the state shaped the development of distinctive secular settlements in the United States and Australia in the late nineteenth and early twentieth centuries. But it also suggests that paying closer attention to how the state structures secularization can have payoffs in explaining the origins and consequences of the diverse array of secular settlements that we see in the world today. The applicability of a state-centered approach to secularization may help explain interesting variation in secularization among otherwise similar countries—Spain and Portugal, for instance, or the Netherlands and Germany [Fox Reference Fox2008]. Future research should investigate how variations in the structure of the state might have contributed to the diverse policies and practices associated with these and other cases.
The state profoundly shaped how secularization unfolded in Australian and American education between 1850 and 1950. Explaining why secularization proceeded differently in these two countries requires close attention to the structure of the state. Such an insight has potentially broad applicability to other cases, and secularization scholars would be wise to pay greater attention to the state as they continue to revisit secularization theory.
Acknowledgments
An earlier version of this paper was presented at the annual meetings of the Society for the Scientific Study of Religion in October 2009, and the American Sociological Association in August 2010. I thank Margaret Weir, Marion Fourcade, Phil Gorski, Jerome Baggett, Juan Fernández, Lynne Gerber, John Hall, Kristen Jafflin, Ben Moodie, Sarah Quinn, Dana Williams, Nick Wilson, and members of the Power and Inequalities Workshop at the University of California, Davis, for helpful feedback; and archivists at State Records New South Wales, the Noel Butline Labour Archive and the Uniting Church Archive for assistance locating documents. For generous financial support, I thank the National Science Foundation (Dissertation Improvement Grant #0727814); the University of California Office of the President Pacific Rim Research Program; the Institute of International Studies and Department of Sociology at the University of California, Berkeley; and the Social and Political Theory Program, Research School of Social Sciences, Australian National University.
