âSecularâ as a social category has religious roots and can be understood only in intimate juxtaposition to âthe religious.â For there to be a classificatory option of religion, its âotherâ (the not it) must exist, at least rhetorically.1 Complicating the use of the concept is its application to phenomena of varying scope and abstraction. Most commonly, the secular and its kindred secularism are employed to reference the political dimensions of that which is explicitly not religious. For instance, âsecularâ is often used to categorize the official policies of nation-states with disestablished or nonestablished relationships to religious traditions and organizations, while âsecularismâ is understood as an ideological framing of political projects aimed at achieving or maintaining secular cultural space.2
Here we are primarily concerned with secular individuals rather than institutions. The two interpenetrate, but not necessarily in expected ways, as evinced by the history of American freethought. We use the term âsecularityâ to refer to the meaning-making strategies people employ who have an explicitly anti- or nonreligious referent. All people utilize an understanding of meaning that is not explicitly religious in some, indeed many, circumstances; but we are interested in individuals for whom ultimate questions of reality, meaning, and identity are not religiousâat least along some basic, traditional dimensions. Our interest is in individuals whose primary identities, experiential perceptions, and routinely utilized cultural frameworks are nonreligious in nature.3
But what makes a person âsecularâ? This is a straightforward yet deceptively difficult question. In asking what counts as an instance of secularity, we are confronted with theoretical baggage freighted by the term. Two issues are of particular concern in this regard. First, the wide-ranging impact and historical importance of secularization theories have produced substantial confusion along with polemical connotationsâboth political and intellectual.4 In short, protracted conceptual and polemic debates have hindered the development of an empirical account of secularity. In analyzing patterns and trends in American secularity, we do not endorse secularism as right, true, or inevitable with the advance of time and culture. Conversely, we also do not endorse the conflation between the âreligiousâ and the âgoodâ that occurs within ethical frameworks laid out in religious and popular (and also sometimes academic) discourse. The notion that the nonreligious are immoral, depraved, or evil is clearly polemical rather than empirical.5 Rather, we see the moral value of religion in individualsâ lives as highly variable and view questions about the social impact of secularity and religiosity as empirical rather than normative issues. We confer neither denigration nor veneration, blessing nor curse. These are judgments for others to make, and they will undoubtedly continue to do so with the passionate energy that accompanies perceived righteousness. Such proclamations are not of interest to us, except as data.
Beyond the moral and polemical connotations entwined with the use of religious and secular categories, their binary classification imposes an unnecessarily strict delineation. Imposing bifurcation onto complex social realities forces a loss of accuracy, especially concerning cases that fall near the borderline. Of course, there will always be conceptual ambiguity bordering the edges of any classification scheme used to simplify a diverse reality, but a binary understanding of secular/religious is too simple and is misleading for the development of a fuller understanding of secularities as cultural phenomena.6 To move beyond a binary classification requires the consideration of multiple dimensions of secularity and religiosity.
Basic Classifications
We use the inverse of the following dimensions of religiosity as keys to classifying secularity: affiliation, belief, and practice.7 For affiliation, we examine whether someone claims any religious affiliation when the option of âno religionâ or ânoneâ is specifically givenâin other words, whether a person claims no affiliation with an established religion. Claiming no religion does not necessarily mean that people consider themselves anti- or nonreligious in their private lives, merely that they do not identify with any organized, public religion. Religious ânonesâ may well be nonreligious privately too, but this is an open question. We define the category as:
Religiously NonaffiliatedâIndividuals who claim no affiliation with an organized religion.
The second dimension we use is theistic belief. Here we are primarily interested in two broad categories of nontheism, atheism and agnosticism. We define atheism in a manner consistent with its popular use, meaning people who do not believe in god.8 Further distinctions can be drawn within this category, such as between âhardâ atheists (who claim definitive proof for the lack of godâs existence) and âsoftâ atheists (who do not believe, but also do not claim definitive proof). Although these distinctions are meaningful for the way people rationalize atheism, we focus on the more general category. We consider agnostics to be those who believe knowledge of godâs existence or nonexistence is beyond human capacity. For example, an agnostic might answer a survey question about the existence of god by selecting an option such as âI donât know and there is no way to find out.â For our purposes, we define those with theistic dis- and nonbelief as:
(1) AtheistsâIndividuals who do not believe theistic claims.
(2) AgnosticsâIndividuals who assert that theistic claims are unverifiable in principle.
We refer to atheists and agnostics as âdisbelieversâ and ânonbelievers,â respectively.9 These categories are defined in reference to questions of theism, understood in the predominantly monotheistic manner of the Abrahamic traditions. To be clear, atheism and agnosticism do not always mean complete dis- or nonbelief in all supernatural entities, as someone may be agnostic about the existence of god but certain about the existence of ghosts.10 Although clearly not true of the majority of atheists or agnostics, this possibility allows us to define the categories in reference to the most culturally central question of supernaturalism in the U.S., then employ the categories to examine the relative absence or presence of other supernatural beliefs.
We also use a category produced by considering those persons who have no religious affiliation and theistic belief in conjunction. We consider theistic dis- or nonbelief to be the most salient marker of oneâs secular identity. That is, self-identifying as someone who does not believe in god is a more prominent marker of identity than saying one is not affiliated with an organized religion.11 If individuals respond to a query about theism by identifying with atheism or agnosticism, we use those designations as the starting point for further examining other social dimensions of their lives; however, theistic dis- or nonbelief and claiming no religion need not necessarily occur in conjunction. A person may quite reasonably be a theistic doubter but value other aspects of participating in a religious community such as camaraderie, social support, ethnic identification, or familial obligations. Such âbelonging without believingâ is relatively common in Scandinavian countries with established national religious traditions. In the U.S., only 1% of the overall pooled 1972â2010 waves of the General Social Surveys (GSS) fit the designation of belonging without believing. Compared to atheists and agnostics who also claim no religious affiliation, those who are affiliated nonbelievers tend to be older, make more money, have more children, and are slightly more likely to be politically conservative. Not surprisingly, affiliated nonbelievers also attend religious services and pray more than nonaffiliated nonbelievers. People fitting this classification according to GSS data were also disproportionately Jewish. Although theistic dis- and nonbelievers who maintain at least symbolic affiliation with a religious tradition certainly warrant inquiry, for the sake of parsimony we focus on nonaffiliated believers and theistic dis- and nonbelief as three mutually exclusive categories of secularity.
Conversely, individuals may not be affiliated with an organized religion but maintain some form of theistic belief. This trend is so prevalent in Western Europe that it has inspired a debate about the meaning and significance of âbelieving without belonging.â12 As we will see, the privatization of religious belief has also become increasingly common in the United States. Accordingly, we distinguish nonaffiliated believers as a useful category in an American context:
(3) Nonaffiliated BelieversâIndividuals who claim no religious affiliation but maintain some form of theistic belief.
A final category of secularity we distinguish is people who claim both religious affiliation and theistic belief, but do not engage in religious practice with any relative frequency. Individuals in this category can be considered as having primarily symbolic religious identities; that is, they maintain a self-understanding that includes religious elements, but they do not engage the ritual dimensions of religion in either public or private. Sociologist N. J. Demerath has termed this phenomenon âcultural religion,â meaning that religious symbols retain some emotive and cultural power for individuals even as they disengage from actively practicing religious communities.13 Cultural religion is a more ambiguous delineation than those of self-identification. Unlike religious ânonesâ or theistic nonbelievers, the culturally religious self-identify as religious and we as researchers determine where to draw the line separating the nonpracticing from the practicing.
To capture whether individuals are disengaged from religious practice, we primarily use two measures of religiosity, one public and one private. To assess public expression of religious commitment, we use attendance at religious services. For private religious practice, we use questions that ask about prayer outside of religious services. For each dimension, we allow individuals to engage in religious practice minimally and still be classified as culturally religious. We consider attendance at religious services twice a year or less to be evidence of a lack of public practice. That is, we consider âChristmas and Easterâ Christians or Jews who attend temple annually on a high holiday but do not pray to be culturally religious rather than actively religious. Thus, our final category is:
(4) Culturally ReligiousâIndividuals who claim religious affiliation and theistic belief, but rarely (if ever) attend religious services or pray privately.
The categorization of individuals as culturally religious is dependent upon the response options available in a given survey. Our pragmatic rule of thumb for classification is to consider the lowest options on a practice question to be nonpracticing; however, response options vary by survey, making the delineation more practical than ideal. For example, there are different response options on religiosity questions in our two primary data sources. For the Pew Religious Landscape Survey, applicable respondents who selected ârarelyâ or âneverâ for both public (religious service attendance) and private religious expression (prayer outside religious services) are classified as culturally religious. For the General Social Surveys, we classify applicable respondents as culturally religious if they answer that they pray outside of religious services less than once a week and attend religious services once a year or less. Such categorization is clearly debatable. Some might well argue for either more stringent or looser conceptualization, but overall it reasonably outlines the category of interest without being so rigid as to leave only a tiny grouping.14 For the most part, we use the four mutually exclusive categories of atheist, agnostic, nonaffiliated believer, and culturally religious as a typology for understanding both the diversity and unity of American secularism. Those who do not fit one of these four categories are classified as âactively religious.â
At times, we also distinguish between individuals raised within a religion who dropped out and those who were raised and stayed outside of religion.
ApostatesâIndividuals who consider themselves to have been religiously affiliated as children, but who are religiously unaffiliated as adults.
Socialized SecularsâIndividuals who consider themselves to have been secular as children, and who remain religiously nonaffiliated as adults.
We use this distinction primarily to track changes and trends in overall religious/secular composition, and also to examine how socialization affects the resulting expressions of secularity.
These categories highlight the variety of secular expressions, while also not slicing the proverbial pie so thin as to lose utility. At the highest level of specificity, there are as many distinct forms of secularity as there are secularists. All people, religious or otherwise, represent idiosyncratic mixtures of socialization, experiences, and self-understanding. At the opposite end of this continuum, religionists often stereotype secularists as âall the same.â Many staunch secularists paint religious people with similarly broad strokes. Such characterizations are generally negative in intent and effect. By outlining general but distinguishable differences in secular expressions, we attempt to strike a balance between the general and the specific, and in doing so highlight the diversity of secularism(s).15
Our overarching analytic strategy is to compare these groups to one another, as well as to religiously affiliated, theistic-believing, ritually practicing Americans. On occasion, such as when estimating long-term trends or examining groups that are relatively small proportions of our primary samples, we encounter data limitations that force us to use the rougher classification of religiously affiliated versus nonaffiliated. We do so only when there is no recourse. A more thorough and accurate understanding of secularism requires greater conceptual and methodological refinement than is typically employed, either in popular or academic discourse.