The Spiritual Turn and the Decline of Tradition: The Spread of Post-Christian Spirituality in 14 Western Countries, 1981–2000


should be addressed to Dick Houtman, Department of Sociology, Faculty of Social Sciences, Erasmus University, P.O. Box 1738, 3000 DR Rotterdam, The Netherlands. E-mail:


This article uses data from the World Values Survey to study the spread of post-Christian spirituality (“New Age”) in 14 Western countries (1981–2000, N = 61,352). It demonstrates that this type of spirituality, characterized by a sacralization of the self, has become more widespread during the period 1981–2000 in most of these countries. It has advanced farthest in France, Great Britain, the Netherlands, and Sweden. This spiritual turn proves a byproduct of the decline of traditional moral values and hence driven by cohort replacement. Spirituality's popularity among the well educated also emerges from the latter's low levels of traditionalism. These findings confirm the theory of detraditionalization, according to which a weakening of the grip of tradition on individual selves stimulates a spiritual turn to the deeper layers of the self.


The debate about religion, modernity, and secularization is livelier than ever since secularization theory, once sociology of religion's proud theoretical flagship, has run into stormy weather. Once considered an empirically sound theory by the social-scientific community, many now feel that it has been exposed as a mere ideology or wish dream, intimately tied to the rationalist discourse of modernity (e.g., Hadden 1987; Stark and Finke 2000). Even one of its most prominent former spokesmen, Peter Berger (1967), has drastically revised his former position: “The assumption that we live in a secularized world is false. The world today … is as furiously religious as it ever was, and in some places more so than ever. This means that a whole body of literature by historians and social scientists … is essentially mistaken” (Berger 1999:2).

Because today's rapid globalization of Islam and the evangelical Christian upsurge, especially in Africa, Latin America, and East Asia, fly in the face of the expectation that religion is doomed, Berger now considers the decline of the churches in many Western European countries the exception rather than the rule (1999:10). And, indeed, it is hardly contested that church membership, adherence to traditional Christian doctrines, and participation in church rituals relating to birth, marriage, and death have all declined considerably in those countries (e.g., Norris and Inglehart 2004). Precisely because of the one-sided attention to those processes of religious decline, however, the extent to which “new” or “alternative” types of religion blossom outside the traditional Christian realm is still quite unclear (Luckmann 2003; Knoblauch 2003; Stark et al. 2005).

Therefore, in the current article, we attempt to map and explain this “post-Christian” type of spirituality—i.e., not the spiritually inclined churched Christianity as discussed by Roof (1999) and Wuthnow (2003) for the United States, but spirituality standing on its own two feet and broken from the moorings of the Christian tradition. Indeed, in the United Kingdom nowadays, very little overlap proves to exist between the congregational domain and the spiritual milieu (Heelas et al. 2005; Heelas forthcoming). This contemporary type of spirituality emerged in the counterculture of the 1960s and became a core element of the “New Age” movement of the 1980s, disembedding from its origins at the countercultural fringe in the process (Van Otterloo 1999). Sutcliffe and Bowman (2000:11) even remark that “contrary to predictions that New Age would go mainstream, now it's as if the mainstream is going New Age.” Our aim in this article is to study whether post-Christian spirituality has indeed become more widespread during the last few decades and, if so, how this process can be explained. Unlike Heelas et al. (2005), who address a similar question for the United Kingdom, we do not focus on participation in the spiritual milieu, but on beliefs and self-designations that characterize post-Christian spirituality among the general populations of late-modern Western countries.

Theorizing the Spiritual Turn

Characterizing Post-Christian Spirituality

In most of the sociological literature, the concept of spirituality—or “New Age,” to the extent this label is still applied today—is used to refer to an apparently incoherent collection of ideas and practices. Most participants in the spiritual milieu, it is generally argued, draw upon multiple traditions, styles, and ideas simultaneously, combining them into idiosyncratic packages. Spirituality is thus referred to as “do-it-yourself religion” (Baerveldt 1996), “pick-and-mix religion” (Hamilton 2000), a “spiritual supermarket” (Lyon 2000), or “religious consumption à la carte” (Possamai 2003). Possamai (2003:40) even states that we are dealing with an “eclectic—if not kleptomaniac—process … with no clear reference to an external or ‘deeper’ reality.”

Notwithstanding their further differences of opinion, defenders of secularization theory and New Age apologetics also agree on the fragmented character of contemporary spirituality. The former typically deny its social significance by invoking the image of a veritable implosion of religion and consumption, suggesting that contemporary spirituality differs dramatically from traditional types of religion in this respect. Steve Bruce (2002:105) writes, for instance:1

The New Age is eclectic to an unprecedented degree and … is … dominated by the principle that the sovereign consumer will decide what to believe … . I cannot see how a shared faith can be created from a low-salience world of pick-and-mix religion.

New Age apologetics for their part tend to emphasize the spiritual supermarket's unlimited diversity so as to underscore the spiritual milieu's openness to diversity, stressing its seemingly unprecedented opportunities for individual choice and liberty, and characterizing the Christian churches as dogmatic and authoritarian in the process. Illustrative in this respect is the following remark by a trainer in a New Age center:2

New Age is like a religious supermarket. All aspects of religion … are put together on a heap and people can choose what is best for them at that moment in time. And that's the good thing about the New Age world—that nobody claims to have a monopoly on wisdom. Whereas the old religions say “We possess the absolute truth” and “This is the only way to God”, we argue: “There are ten thousand ways” and “There are as many ways as there are people.”

Although, to be sure, neither of those two positions is completely mistaken, both overestimate the fragmented character of post-Christian spirituality. True, today's spiritual consumers sample their personal diets from the well-packed shelves of the late-modern spiritual supermarket, but underneath this diversity lies a commonly held belief that has all too often been neglected. This shared doctrine not only provides a substantial explanation for the bewildering diversity of the spiritual milieu, but also contradicts the alleged “authenticity” or “individualism” of New Agers. As it happens: “The great refrain, running throughout the New Age, is that we malfunction because we have been indoctrinated … by mainstream society and culture” (Heelas 1996:18). The latter are thus conceived of as basically alienating forces, held to estrange one from one's “authentic,”“natural,” or “real” self—from who one “really” or “at deepest” is:

Perfection can be found only by moving beyond the socialized self—widely known as the “ego” but also as the “lower self”, “intellect” or the “mind”—thereby encountering a new realm of being. It is what we are by nature. Indeed, the most pervasive and significant aspect of the lingua franca of the New Age is that the person is, in essence, spiritual. To experience the “Self” itself is to experience …“inner spirituality”. … The inner realm, and the inner realm alone, is held to serve as the source of authentic vitality, creativity, love, tranquility, wisdom, power, authority and all those other qualities that are held to comprise the perfect life. (Heelas 1996:19, italics in original)

This, then, is the main tenet of post-Christian spirituality: the belief that in the deepest layers of the self the “divine spark”—to borrow a term from ancient Gnosticism—is still smoldering, waiting to be stirred up, and succeed the socialized self. Getting in touch with this “true,”“deeper,” or “divine” self is not considered a “quick fix,” but rather understood as a long-term process: “‘Personal growth,’” Hanegraaff rightly notes, “can be understood as the shape ‘religious salvation’ takes in the New Age Movement” (1996:46). Reestablishing the contact with the divine self is held to enable one to reconnect to a sacred realm that holistically connects “everything” and to thus overcome one's state of alienation. True spiritual evolution even transcends the boundaries of this life because New Agers generally believe in reincarnation (e.g., Hanegraaff 1996; Heelas 1996; Rose 2005). Paradoxically, then, New Agers believe that working on the self raises consciousness about the true, divine nature of the world as a whole—it leads to the acknowledgment that “[a]ll life—all existence—is the manifestation of Spirit” (Bloom, quoted in Rose 2005:31).

Post-Christian spirituality, in short, constitutes a basically romanticist conception of the self that is intrinsically connected to an immanent conception of the sacred. It “lays central stress on unseen, even sacred forces that dwell within the person, forces that give life and relationships their significance” (Gergen 1991:19). This conviction differs considerably from the traditional Christian belief that “[t]he truth is ‘out there’ rather than within,” that “the divine is transcendent rather than immanent” (Heelas et al. 2005:22). Post-Christian spirituality rejects, at the same time, the premise of secular rationalism that “truth—if attainable at all—can only be discovered by making use of the human rational faculties” (Hanegraaff 1996:519). As such, post-Christian spirituality entails an epistemological third way of “gnosis,” rejecting both religious faith and scientific reason as vehicles of truth. Rather, it is held that one should be faithful to one's “inner voice” and trust one's “intuition”:

According to [gnosis] truth can only be found by personal, inner revelation, insight, or “enlightenment.” Truth can only be personally experienced: in contrast with the knowledge of reason or faith, it is in principle not generally accessible. This “inner knowing” cannot be transmitted by discursive language (this would reduce it to rational knowledge). Nor can it be the subject of faith… because there is in the last resort no other authority than personal, inner experience. (Hanegraaff 1996:519, italics in original)

As an heir of the long-standing esoterical tradition, post-Christian spirituality can thus be understood as an alternative “third option” in Western culture, informed by culture criticism (Hanegraaff 1996:517–18). Those involved do not pursue meaning and identity from “pregiven” authoritative sources, located outside the self (e.g., the answers offered by science and the Christian churches), but want to rely on an “internal” source, located in the self's deeper layers. It is this “dogma of self spirituality” that not only accounts for the much emphasized diversity at the surface of the spiritual milieu—an inevitable outcome when people feel that they need to follow their personal paths and explore what works for them personally—but that simultaneously provides it with unity at a deeper level.3 Post-Christian spirituality can be characterized, in short, by the idea that the self is divine and by the immanent conception of the sacred that goes along with this.

Explaining the Spiritual Turn

No such thing as “the” theory of secularization exists (Tschannen 1991). Casanova (1994:7) distinguishes three different propositions, conceiving of secularization as differentiation, privatization, and religious decline, respectively, with “the thesis of the differentiation of the religious and the secular spheres [as] the still defensible core of the theory of secularization.” And, indeed, notwithstanding their further disagreements, defenders of secularization theory like Wilson (1976, 1982) and critics like Luckmann (1967, 2003) agree that a process of institutional differentiation has taken place in that social functions have increasingly come to be dealt with by specialized institutions. Medieval art, for instance, was basically religious art; during the Renaissance science and religion were still inextricably intertwined; and today's division of state and church in Western countries constitutes the outcome of a long and painful historical process (e.g., Wilson 1982; Luckmann 2003).

With the institutional separation of the economy, the family, the state, science, art, etc., these realms increasingly came to be governed by their own particular institutional logics (compare Bell 1976). Examples are the principles of caring and nurturing in the family, of maximization of utility in the economy, of pursuing truth in science, and beauty in art. As a consequence, religion lost its ability to morally overarch all of society as a sort of “sacred canopy” (Berger 1967) and the religious domain became one among many, significantly affecting its impact on politics, the economy, art, and science. Due to this disintegration of the sacred canopy, people came to find themselves confronted with a value pluralism that erodes the unquestioned legitimacy of the traditional moral values bound up with the Christian tradition.

The resulting process of “detraditionalization” (Heelas 1995) or “individualization” (Beck and Beck-Gernsheim 2002) is not simply the aggregated result of individual desires or choices for more liberty—“Individualization is a fate, not a choice,” as Bauman (2002: xvi) aptly remarks. As external and authoritative sources of meaning and identity lose their grip on individuals, the range of biographical and life-styles choices nevertheless widens considerably:

It is …[the] level of preconscious “collective habitualizations,” of matters taken for granted, that is breaking down into a cloud of possibilities to be thought about and negotiated. The deep layer of foreclosed decisions is being forced up into the level of decision making. (Beck and Beck-Gernsheim 2002:6)

Although the foregoing is by and large agreed upon, radically different theoretical positions can be distinguished as far as the implications for individual religiosity are concerned. First, the “decline of religion thesis, often misinterpreted as “the” secularization theory (Casanova 1994:25–35), assumes that Christian religiosity and traditional moral values give way to a rationalist worldview. Bryan Wilson, for instance, writes:

In contemporary society, the young come to regard morality—any system of ethical norms—as somewhat old-fashioned. For many young people, problems of any kind have technical and rational solutions. (Wilson 1982:136; see also Dobbelaere 1993:15)

There are embarrassingly few studies that systematically map the worldviews of the unchurched, however, and research even suggests that the assumption that secularization and detraditionalization give rise to a rationalist worldview may simply be mistaken:

A diminishing faith in rationality and a diminishing confidence that science and technology will help solve humanity's problems … has advanced farthest in the economically and technologically most advanced societies. (Inglehart 1997:79)

Be this as it may, our ambition in this article is to explain a spiritual rather than a rationalist turn. At this point an alternative theoretical logic comes in, which agrees that the erosion of traditional moral values loosens the grip of external and authoritative sources of meaning and identity, but assumes this to stimulate a spiritual rather than a rationalist turn. The spread of “elective biographies,”“reflexive biographies,” or “do-it-yourself biographies” (Beck and Beck-Gernsheim 2002:3) entails “precarious freedoms” (2002:16), after all. In dealing with their newly acquired liberty, late-modern individuals are thrown back upon themselves (Heelas 1995:2) and easily experience this as a burden.

Robbed of the protective cloak of “pregiven” or “self-evident” meaning and identity, the late-modern condition conjures up nagging questions that haunt the late-modern self: “What is it that I really want?”“Is this really the sort of life I want to live?”“What sort of person am I, really?” Because only one's personal feelings, emotions, intuitions, and experiences remain as touchstones for meaning and identity, one embarks on a voyage of discovery to the deeper layers of the self to dig for “real” meaning and “real” identity there. And as we have seen above, this is precisely the key tenet of post-Christian spirituality: the conviction that meaning and identity can only be derived from an “internal” source, located in the deeper layers of the self. We suggest, in short, that post-Christian spirituality becomes more widespread if traditional moral values decline, with those individuals who reject these values most strongly being most receptive to post-Christian spirituality.


The theory of detraditionalization outlined above suggests that post-Christian spirituality becomes more widespread if adherence to traditional moral values declines. Because it is hardly contested in the literature that such a process of detraditionalization has actually taken place—it has in fact been demonstrated by Inglehart (1997)—we expect that post-Christian spirituality has become more widespread during the last few decades (Hypothesis 1). Our theory of detraditionalization further suggests that this spread is driven by a process of cohort replacement—the gradual replacement of less spiritually inclined older birth cohorts with more spiritually inclined younger ones (Hypothesis 2). If the latter hypothesis is confirmed, the well-known circumstance that New Agers tend to be younger than the population at large (e.g., Stark and Bainbridge 1985; Becker et al. 1997; Houtman and Mascini 2002) is not simply a consequence of their relative youth (a “life-cycle effect”), but rather of their having been born later (a “cohort effect”), hence pointing out a process of historical change.

Obviously, our theory also predicts that in particular “posttraditionalists,” i.e., those who reject traditional moral values, are likely to sacralize their selves (Hypothesis 3). This hypothesis is consistent with the well-established finding that post-Christian spirituality is not only typical of the young, but of the well educated, too (e.g., Stark and Bainbridge 1985; Becker et al. 1997; Houtman and Mascini 2002), because precisely these two demographic categories stand out as the least traditional ones as far as their moral values are concerned.4 Our theory of detraditionalization also suggests, in short, that the younger age cohorts and the well educated display highest levels of post-Christian spirituality because of their high levels of posttraditionalism (Hypothesis 4 and Hypothesis 5, respectively).5

Given our reliance on survey data, there is of course no way to prove that “posttraditionalism”—referred to as “moral individualism” in an earlier article (Houtman and Mascini 2002)—is actually the “cause” of post-Christian spirituality. The relationship may in fact be reverse (or, more likely, working in both directions). Although we treat post-Christian spirituality as the dependent and posttraditionalism as an independent variable in our statistical analyses, we hence have no intention of making bold causal claims.6 The principal idea we want to test is simply whether detraditionalization and the spiritual turn are indeed two intimately connected processes. To convey our causal modesty to the reader, we have tried to prevent overly “mechanistic” and “causal” language as much as possible.

Data and Measurement

World Values Survey (1981–2000)

Studies about post-Christian spirituality are typically based on qualitative research, employing semistructured interviews, ethnography, case studies, content analysis, etc. Those studies do not permit a systematic comparison of countries and periods, so as to find out whether post-Christian spirituality has indeed become more widespread within a particular country, whether the same applies to other countries, and in which countries it has expanded most. It is this type of questions, addressed in the current article, which requires survey data and quantitative research methods.

Although good scales for the measurement of post-Christian spirituality have become available during the last few years (e.g., Granqvist and Hagekull 2001; Houtman and Mascini 2002), such scales are unfortunately absent from the large international survey programs that enable comparisons across countries and across time. The World Values Survey is no exception to this general rule and hence precludes a theoretically sophisticated measurement of the extent to which one identifies with post-Christian spirituality. We nevertheless feel that the World Values Survey can be used to measure post-Christian spirituality in a satisfactory, albeit crude, way by strategically combining answers to multiple questions. And whereas no better data sources are available to satisfy our research needs, we feel that some pragmatism is justified—especially so because the data of the World Values Survey are otherwise perfectly suited for our purposes. This is so for three reasons.

First, the three available waves of the World Values Survey (WVS) (1981, 1990, 2000) cover a range of 20 years. Of course, one would prefer to also have comparable data for 1970, or even earlier. But then again, 20 years is quite an impressive timespan, especially if we realize that the large surveys fielded today hardly include better measures for post-Christian spirituality. Moreover, it is often argued that the expansion of post-Christian spirituality has particularly taken place during the 1980s (e.g., York 1995; Hanegraaff 1996), after its first emergence in the 1960s and 1970s counterculture (e.g., Roszak 1968; Zijderveld 1970; Bell 1976). Second, the WVS covers a substantial number of countries. Obviously, not all of those have been included in all three rounds of data collection, not all of those are Western countries with a Christian heritage, and the crucial questions have not always been asked. Nevertheless, for 14 Western countries with a Christian heritage, we have sufficient data to map and explain the spiritual turn in late modernity: France, Great Britain, West-Germany, Italy, Netherlands, Denmark, Belgium, Spain, Ireland, the United States, Canada, Norway, Sweden, and Iceland.7 A third reason why the WVS perfectly meets our needs is that it strongly emphasizes the measurement of adherence to traditional moral values. This makes it highly useful for testing our theory on why a spiritual turn has occurred in the first place, as we will explain below.

All samples are representative cross-sections of the respective adult populations (18 years or older in three-quarters of the country-year combinations; 15, 16, or 17 or older in the remaining quarter). The number of respondents averages 1,461, ranging from 702 (Iceland, 1990) to 4,147 (Spain, 1990), and the average age across all country-year combinations is 43 years. For details about sampling and data collection, the reader is referred to the technical reports of the WVS waves of 1981 and 1990 (World Values Study Group 1994) and of 2000 (European Values Study Group and World Values Survey Association 2004).

Measuring Post-Christian Spirituality

The questionnaire of the World Values Survey contains one very simple question that more or less directly taps post-Christian spirituality as distinguished from traditional Christian religion. Respondents have been asked which of the following four statements comes closest to their personal beliefs: “there is a personal God,”“there is some sort of spirit or life force,”“I don't really know what to think,” and “I don't really think there is any sort of spirit, God, or life force.” Although “belief in some sort of spirit or life force” does not necessarily imply belief in the immanence of the sacred, previous research has pointed out that those who choose this option score substantially higher on a valid and reliable scale for New Age affinity than those who choose any of the three others (Houtman and Mascini 2002:462–63).

Although the questionnaire contains no other questions that directly tap affinity with post-Christian spirituality, we feel that four additional dichotomous indicators can be constructed by capitalizing on the circumstance that post-Christian spirituality sets itself apart from both the Christian churches and secularist rationalism (e.g., Hanegraaff 1996; Aupers and Houtman 2003).8 Consequently, answers that may crudely indicate post-Christian spirituality, but may also tap less orthodox Christian affinities, can further be polished by combining them with answers that unambiguously reveal that one critically distances oneself from the Christian church. Likewise, answers indicating such a critical distance can be polished by combining them with answers that indicate a rejection of secularist rationalism. The former strategy enables us to demarcate post-Christian spirituality from Christianity; the latter to demarcate a rejection of the Christian churches from secularist rationalism.

The first additional indicator arrived at in this way robs the belief in a life after death of its traditional Christian associations by combining it with the feeling that the churches do not give adequate answers to people's spiritual needs. We take the combination of belief in a life after death and this criticism of the churches to indicate post-Christian spirituality and the three remaining combinations to indicate its absence. The second additional indicator combines a belief in reincarnation (a principal tenet of New Age, closely related to the belief in a divine self, as we have argued above) with an absence of belief in God (compare Heelas 1996:112). Those first two additional indicators solve the awkward problem of demarcating the boundary between post-Christian spirituality and secular-humanist conceptions of “expressive individualism” (e.g., Bellah et al. 1985; on this boundary problem, see Heelas 1996:115–17). Both indicators are precisely convincing because they do not simply tap (arguably secular) self-expression as a key value, but indeed express belief in the existence of a self that is essentially immortal.

The circumstance that post-Christian spirituality presents itself as an alternative for both Christian religion (“faith”) and secularist rationalism (“reason”) is used to construct two further additional indicators. Both capture the idea of post-Christian spirituality as a third way beyond faith and reason. The first has been constructed by cross-tabulating whether one considers oneself a convinced atheist (reason) and whether one belongs to a religious denomination (faith). We take a rejection of both of those identities (not belonging to a religious denomination, but not considering oneself a convinced atheist either) to indicate post-Christian spirituality, coding the three remaining categories as its absence. Likewise, we conceive of having no or not very much confidence in the churches, although not considering oneself a convinced atheist, as a final indicator for post-Christian spirituality. Table 1 displays the five resulting indicators.

Table 1. 
Indicators% No Affinity% Affinity% Valid
• Believes in the existence of a spirit or life force.
• Believes in a life after death, but thinks that the churches do not give adequate answers to people's spiritual needs.83.716.369.9
• Believes in reincarnation, but does not believe in God.97.0 3.069.9
• Does not belong to a religious denomination, but does not consider oneself a convinced atheist either.
• Does not consider oneself a convinced atheist, but has not very much or no confidence in the churches.58.941.189.7

More than 40 percent consider themselves neither convinced atheists, nor have confidence in the churches. As such, this indicator generates the highest level of post-Christian spirituality. At the other end, a mere 3 percent believe in reincarnation but do not believe in God. The three remaining indicators take up positions between those two extremes. A principal component analysis yields a first factor that explains 30 percent of the variance and a second factor that explains 21 percent (hence, eigenvalues for the first two factors are 1.52 and 1.05, respectively, while those for the three remaining ones are 0.90, 0.85, and 0.69, respectively). Because the rotated two-factor solution is difficult to interpret with one of the indicators loading on both factors, we choose the unrotated first factor instead.9 The two lowest loadings on this unrotated first factor are both 0.44 (for belief in a spirit or life force and for belief in a life after death, while feeling that the churches do not give adequate answers to people's spiritual needs), while all three others are higher than 0.50.

Although Cronbach's alpha is not higher than 0.42—hardly surprising, given that we are dealing with merely five crude dichotomous indicators—and cannot be improved by excluding any of the indicators, all zero-order correlations are positive and significant (p < 0.001, two-sided test) and corrected item-total correlations range from 0.17 to 0.30. These findings obviously do not indicate that a linear combination of our five dichotomous indicators constitutes a reliable measure of post-Christian spirituality. Neither do we claim that it constitutes a theoretically sophisticated measurement. What we do claim, however, is that an index based on these five indicators can nevertheless be used to crudely map and explain the spiritual turn in 14 Western countries since 1981.

All five dichotomous indicators (1: post-Christian spirituality; 0: no post-Christian spirituality) have first been standardized and the resulting z-scores have next been combined into an index ranging from 0 to 10. Correcting for the number of valid scores, scores have been assigned to all respondents with valid scores on at least three of those five indicators. For one of the 42-year-country combinations—Norway, 2000—scores could not be assigned due to missing values. This produces measurements of affinity with post-Christian spirituality for 92 percent of the 61,352 respondents.

Other Measures

Following Houtman (2003), posttraditionalism has been operationalized by combining three measures that tap acceptance or rejection of traditional moral values (i.e., acceptance or rejection of the traditional hierarchical relationship between parents and children, of traditional values pertaining to sexuality, and of traditional male and female gender roles) with Inglehart's index for postmaterialism.

First, self-direction or conformity as a parental value has been measured by means of respondents' selection of a maximum of five from a list of qualities that may be encouraged in children. Their evaluation of six of those qualities is used here: “determination/perseverance,”“imagination,” and “independence” (indicating “self-direction”) and “obedience,”“religious faith,” and “good manners” (indicating “conformity”). With all of those qualities coded either 0 (“not chosen”) or 1 (“chosen”), the more goals one selects from the former three and the fewer from the latter three, the higher one's level of posttraditionalism.10

Second, sexual permissiveness has been measured as judgments about the acceptability of five activities: married men/women having an affair, sex under the legal age of consent, homosexuality, prostitution, and abortion. Scores range from “never justified” (1) to “always justified” (10) and high scores are taken to indicate high levels of posttraditionalism.11

Third, three questions have been used to measure the degree to which traditional gender roles are accepted or rejected. The first two are statements that “a woman has to have children to be fulfilled” and that “a single woman should have the right to have a child” and the third is a question about whether one feels that marriage is an out-of-date institution.12

Fourth and finally, Inglehart's index for postmaterialism is constructed on the basis of the prioritization of four political goals by the respondents. Those who select “Giving the people more say in important government decisions” and “Protecting free speech” as the two most important goals are coded “postmaterialists,” and those who choose “Maintaining order in the nation” and “Fighting rising prices” are coded “materialists.” Remaining respondents are coded as a mixed category in between.13

As expected, those four measures are strongly related among themselves. A second-order factor analysis produces a first factor that accounts for 46 percent of the variance, with factor loadings of 0.65 (rejection of traditional gender roles), 0.59 (postmaterialism), 0.68 (emphasis on self-direction rather than conformity as a parental value), and 0.77 (sexual permissiveness). Scores for posttraditionalism are therefore assigned as mean standardized scores to 98.7 percent of the respondents with valid scores on at least three of those four measures. In effect, those who are postmaterialists, critical of traditional gender roles, critical of conformity as a parental value, and sexually permissive receive the highest scores on posttraditionalism.

The three final variables to be included in our analysis are operationalized fairly straightforwardly: age is measured in years, birth year as the year in which one was born, and education as the age at which one completed full-time education (coded into 10 categories ranging from 1, younger than 13 years, through 10, older than 20 years).

Mapping and Explaining the Spiritual Turn

Mapping the Spiritual Turn

Table 2 displays the distribution of post-Christian spirituality across the 14 countries and three survey years. Italy, Canada, and Iceland prove exceptions to the general pattern of change that emerges, in that post-Christian spirituality has declined in those three countries since 1981. With the exception of those three countries, however, the general pattern of change is clear enough. Measured by either Cohen's d—an estimate of the effect size that ranges from 0 to 1—or Pearson's correlation between year and individual-level spirituality, post-Christian spirituality has become more widespread since 1981 in the 11 remaining countries. Although the increase is certainly not strong (a Cohen's d of 0.20 is generally taken as the boundary between a “weak” and a “modest” effect), the overall trend nevertheless confirms that a spiritual turn has been taking place during the last two decades, which confirms our first hypothesis.

Table 2. 
Country1981 M (SD)1990 M (SD)2000 M (SD)Change 1981–2000
Cohen's daPearson's rb
  1. aCohen's d estimates the effect size and is computed as d = (M2000− M1981)/√[(σ2000219812)/2] (Cohen 1988; Rosnow and Rosenthal 1996).

  2. bPearson's r has been computed as the correlation between year (i.e., 1981, 1990, or 2000) and individual-level post-Christian spirituality with two-tailed testing for significance.

  3. *p < 0.05; **p < 0.01; ***p < 0.001.

France1.61 (1.56)1.97 (1.80)2.07 (1.80) 0.27 0.11***
Great Britain1.83 (1.50)2.33 (1.68)1.97 (1.47) 0.09 0.04**
West Germany1.57 (1.21)1.87 (1.43)1.70 (1.37) 0.18 0.03*
Italy1.74 (1.29)1.46 (1.47)1.37 (1.42)−0.27−0.10***
Netherlands2.26 (1.73)2.76 (1.99)3.03 (2.07) 0.40 0.16***
Denmark1.43 (1.32)1.71 (1.51)1.62 (1.41) 0.14 0.05**
Belgium1.35 (1.41)1.69 (1.44)2.10 (1.68) 0.48 0.17***
Spain1.51 (1.34)1.67 (1.32)1.86 (1.52) 0.24 0.09***
Ireland1.02 (1.09)1.27 (1.09)1.53 (1.32) 0.42 0.18***
United States1.04 (1.02)1.44 (1.31)1.57 (1.68) 0.38 0.17***
Canada1.89 (1.44)1.89 (1.69)1.71 (1.87)−0.11−0.04**
Norway1.53 (1.22)1.77 (1.45) 0.18 0.09***
Sweden1.78 (1.23)1.84 (1.53)2.29 (1.63) 0.35 0.14***
Iceland1.88 (1.32)1.53 (1.42)1.65 (1.59)−0.16−0.06***
Total1.55 (1.37)1.77 (1.52)1.86 (1.67) 0.20 0.08***

Post-Christian spirituality has expanded most in the Netherlands, Belgium, Ireland, and the United States. In the latter two countries, despite this growth, it was least widespread in 1981 and those two countries still lag behind today (only for Italy, in which post-Christian spirituality has declined since 1981, lower levels of affinity are found in 2000). The countries that lead the way with highest levels of affinity with post-Christian spirituality are France, Great Britain, the Netherlands, and Sweden. It is clear from those findings that post-Christian spirituality has become more widespread since 1981, confirming the occurrence of a spiritual turn.

Cohort Replacement?

Has the spiritual turn across this 20-year period really been caused by an intergenerational replacement of older age cohorts by more spiritually inclined younger ones? To test our second hypothesis that predicts such to be the case, we distinguish age from birth year to disentangle “age effects” from “cohort effects.”14 We use multilevel analysis with maximum likelihood estimation because we are dealing with so-called nested data, i.e., data in which cases defined at a lower level (typically, and also in this case, respondents) are embedded in contexts defined at a higher level (in this case, combinations of countries and years). Multilevel analysis enables one to assess the importance of both levels for explaining a variable defined at the individual level, to find out which contextual and individual variables are (most) consequential and to analyze whether the effects of variables defined at the individual level vary across contexts.

All variables have been standardized so as to produce standardized regression coefficients that enable a straightforward comparison of the strengths of the effects found. In this case, aimed at the explanation of individual-level post-Christian spirituality, the explanatory role of the contextual level proves limited. About 6 percent of the differences at the individual level can be attributed to differences between the years and countries, whereas the remaining 94 percent is caused by individual characteristics, yielding an intra-class correlation of 0.06.15

Table 3 demonstrates that although post-Christian spirituality has not become significantly more widespread during the period 1981–1990, it has become so in 2000 as compared to 1981 (Model 2). The differences between the three years account for about 8 percent of the differences at the contextual level, indicating that the differences between the countries are much larger than those between the years (as mentioned above, the contexts are combinations of countries and years in this case, after all). Affinity with post-Christian spirituality is nevertheless found among those who have been born more recently rather than among those who are younger, and this cohort effect accounts for the higher level of post-Christian spirituality in the most recent period (Model 4 as compared to Model 2). Needless to say, these findings confirm our second hypothesis that the spiritual turn is driven by cohort replacement: younger age cohorts with stronger affinities with post-Christian spirituality have replaced less spiritually inclined older age cohorts.

Table 3. 
  1. *p < 0.05; **p < 0.01; ***p < 0.001.

Contextual level
 1981 (reference category)
 1990 0.063 0.032
 2000 0.085* 0.016
Individual level (fixed effects)
 Year of birth 0.222*0.189***
 Age 0.031 
−2 Log likelihood157,230.3157,226.3154,168.7154,168.3
Variance contextual level0.0630.0580.0590.058
Variance explained contextual level (%)
Variance individual level0.9430.9430.9120.912
Variance explained individual level (%)

Explaining the Spiritual Turn

Our theory suggests that a process of detraditionalization is responsible for the spiritual turn mapped above. To test whether such is indeed the case, we have conducted a second series of multilevel analyses with posttraditionalism as the key additional variable (besides birth year and education).16

Hypothesis 3 is confirmed. As expected, it is an increase of posttraditionalism that underlies the higher levels of post-Christian spirituality in the most recent period (Model 2 in Table 4), with especially those standing out as most posttraditional embracing post-Christian spirituality (Model 4 as compared to Model 2).17 Finally, the affinity with post-Christian spirituality among the well educated is completely accounted for by their posttraditionalism and that among the younger birth cohorts to a substantial extent (Model 4 as compared to Model 3). This means that Hypothesis 4 and Hypothesis 5 are confirmed, too.

Table 4. 
  1. *p < 0.05; **p < 0.01; ***p < 0.001.

Contextual level
 1981 (reference category)
 1990 0.022-0.0030.003
 2000 −0.010−0.067−0.046
 Mean posttraditionalism 0.151***0.134**0.066
Individual level (fixed effects)
 Year of birth 0.166***0.096***
 Education 0.070***0.001
 Posttraditionalism 0.270***
−2 Log likelihood157,230.3157,213.9148,863.3144,077.4
Variance contextual level0.0630.0420.0450.042
Variance explained contextual level (%)0.033.328.633.3
Variance individual level0.9430.9430.9060.852
Variance explained individual level (%)

Conclusion and Debate

What we are witnessing today is not so much a disappearance of religion, but rather a relocation of the sacred. Gradually losing its transcendent character, the sacred becomes more and more conceived of as immanent and residing in the deeper layers of the self. At least in many places, religion is giving way to spirituality, in short (Heelas et al. 2005; see also Partridge 2004). But yet, theoretical controversy about the future of individual religiosity persists. Bruce (2002) forecasts a by-and-large secular future, whereas Heelas et al. (2005) envisage one in which post-Christian spirituality supersedes religion. As we see it, neither of those two predictions of religion's future is completely satisfactory, however, and two issues are of critical importance in pushing the frontiers of knowledge further forward.

Although Heelas et al. (2005) are correct in claiming that religion is giving way to spirituality, they neglect conversions from Christianity to secularism.18 This is quite unfortunate because previous research has pointed out that posttraditionalists are equally likely to embrace post-Christian spirituality as to reject it along with Christian religion, adopting a basically secularist posture in the process (Houtman and Mascini 2002). The explanation of these two radically diverging responses to the condition of posttraditionalism deserves a prominent position on the research agenda. We have demonstrated in this article that detraditionalization goes along with a spiritual turn and our theory assumes that this is because the decline of tradition conjures up problems of meaning and identity that haunt the late-modern self: “What is it that I really want?”“Is this really the sort of life I want to live?”“What sort of person am I, really?” The validity of this assumption about the role of problems of meaning and identity needs to be studied more explicitly in future research. More specifically, it is vital to test the hypothesis that those posttraditionalists who experience problems of meaning and identity are likely to embrace post-Christian spirituality, while those who do not are likely to adopt a basically secularist stance.

A second important issue relates to Bruce's prediction of a secular future. One of his key arguments in this hinges on the alleged “diffuseness” and “social insignificance” of post-Christian spirituality. These lead him to argue that “the transmission of diffuse beliefs is unnecessary and it is impossible” (2002:99), i.e., that no such thing as socialization into post-Christian spirituality exists, causing it to die out inevitably. He strongly emphasizes the radical fragmentation and individualism of the spiritual milieu to hammer this point home. But consider the following example by Bruce himself:

Findhorn, one of Europe's oldest centers of New Age thought and teaching, requires of those who take part in its various forms of group work that they confine their talk to “I statements.” The point of this is to establish that, while each participant has a right to say how he or she feels or thinks, no-one has a right to claim some extra-personal authority for his or her views. (Bruce 2002:83; italics added)

To be sure, this example does much to underscore the radical individualism of the spiritual milieu. But simultaneously, and ironically, it does more than that. It also demonstrates how this very individualism operates as a socially sanctioned obligation of personal authenticity in the spiritual milieu, thus contradicting precisely the diffuseness and social insignificance posited by Bruce. His failure to capture and theorize this ambiguity causes Bruce to overlook the possibility that participants in the spiritual milieu socialize one another into compliance to the ideal of personal authenticity. Socialization may play an important role after all, in short, undermining the claim of post-Christian spirituality's diffuseness and social insignificance, and hence that of its inevitable waning, too.

We are in fact quite skeptical about the validity of the widespread idea of post-Christian spirituality as “fully privatized” and “purely individualistic” (Aupers and Houtman 2006). There is no such thing as a “New Age church,” and “self-spirituality” differs radically from the Christian dogmas, to be sure. But then again, does this mean that post-Christian spirituality is merely individualistic? Obviously, the spiritual practitioners interviewed by Heelas et al. (2005:27) “[t]ime and time again …[reject] the idea that their relationships with their group members or clients have anything to do with prepackaged … ways of transmitting the sacred.” But even if they do not “[tell] their group members or clients what to think, do, believe or feel” (2005:28), they are likely to tell them that they should take their personal feelings seriously, that a one-sided reliance on thinking at the cost of feeling is detrimental, and that one should follow one's heart. And if they do, post-Christian spirituality is socially constructed because people are socialized into a spiritual discourse about the self (e.g., Hammer 2001; Heelas 2006). We consider this another key issue for future research, especially because the overly confident claims about post-Christian spirituality as “fully privatized” and “purely individualistic” contrast strikingly with the virtual absence of empirical studies into the social construction of self-spirituality.


This article benefited from research conducted during the first author's visiting professorship at the Ecole des Hautes Etudes en Sciences Sociales, Marseille, France (November 2003). It is based on keynote lectures delivered at the conferences Religion and Gender (Sociology of Religion Study Group, British Sociological Association, Lancaster University, United Kingdom, April 11–13, 2005) and Una sociologia dellaspiritualità” (Convegno Nazionale della Sezione Sociologia della Religione, Associazione Italiana di Sociologia, Aosta, Italy, November 25–26, 2005). The authors thank Peter Mascini, Peter Achterberg, Sabine Lauderbach, and three anonymous reviewers for their useful comments on previous drafts.


  • 1

    In a similar vein, the late Bryan Wilson argued 30 years ago that the post-Christian cults “represent, in the American phrase, ‘the religion of your choice,’ the highly privatized preference that reduces religion to the significance of pushpin, poetry, or popcorns” (1976:96; see, for another example, Becker et al. 1997).

  • 2

    Interviewed in the context of a previous study (Aupers and Houtman 2003).

  • 3

    See Aupers and Houtman (2006) for a more elaborate critique of the tendency in the academic literature to portray New Age as “ideologically incoherent” and “radically privatized” and its participants as “consumers in a spiritual supermarket.”

  • 4

    To cite the most influential example only, Inglehart (1977, 1990, 1997) demonstrates that “postmaterialism” is especially found among the younger age cohorts and the well educated and is strongly related to posttraditionalism, conceived of as a rejection of traditional values pertaining to gender roles, sexuality, child-rearing, and so forth (1997:47). Studies into the propensity to obey authorities have always arrived at precisely the same conclusion: the young and the well educated are least likely to do so (see Houtman 2003 for a review of the relevant studies).

  • 5

    Why women, like the well educated and the young, also display high levels of affinity with spirituality remains a difficult puzzle to solve (Houtman and Mascini 2002:468) because men and women do not differ with respect to posttraditionalism. Solving this “gender puzzle” (Heelas et al. 2005) requires a gendering of the theory of detraditionalization (see also Woodhead 2005, forthcoming) that we attempt elsewhere (Houtman and Aupers forthcoming).

  • 6

    That “posttraditionalism” and “post-Christian spirituality” are two different manifestations of a single more general phenomenon is a claim that is more difficult to uphold. Whereas previous research has pointed out that post-traditionalism goes together as smoothly with post-Christian spirituality as with nonreligious secularism (Houtman and Mascini 2002; see also the discussion in the concluding section of this article), this argument boils down to the position that the latter is also a manifestation of this same more general phenomenon, implying the awkward claim that there is no basic difference between post-Christian spirituality and nonreligious secularism.

  • 7

    For reunified Germany in the data collection of 2000, we have included only the Länder that used to be part of West Germany in our data analysis.

  • 8

    Following this strategy, we assume that “setting oneself apart from the Christian churches” and “setting oneself apart from secularist rationalism” can be measured by means of the same (combinations of) indicators in the 14 countries. We thus assume, for instance, that belonging or not belonging to a Christian denomination does not have a radically different meaning in countries with state churches and in countries with a variety of different denominations, that considering oneself a convinced atheist does not have a radically different meaning in a country like France and in a country like Italy, etc. Needless to say, whether assumptions such as these are valid is a question that goes way beyond the purposes of this article.

  • 9

    “Belief in reincarnation, but no belief in God” loads on both factors (Factor 1: 0.30; Factor 2: 0.49), “No religious denomination, but not convinced atheist either” and “Not convinced atheist, but not much or no confidence in the churches” both load on the first factor (0.83 and 0.87, respectively), and “Belief in spirit or life force” and “Belief in a life after death, while feeling that the churches do not give adequate answers to people's spiritual needs” both load on the second factor (0.64 and 0.74, respectively). Apart from the problem of interpretation, another reason for choosing the one-dimensional solution is that the two two-item indexes that would result after deleting the first item mentioned above would be no more reliable than the index based on the first unrotated factor (Cronbach's alpha for the index based on the first rotated factor would be 0.42 and for that based on the second one 0.21).

  • 10

    An overall factor analysis produces a single factor with an eigenvalue higher than one, explaining 28 percent of the common variance. All six factor loadings are higher than 0.40 and the signs of the loadings for the indicators for self-direction are positive, whereas those for conformity are negative. If a factor loading for a particular indicator was below 0.30 for a country-year combination, it has been coded missing for that particular country-year combination. After reversing the scores on the indicators tapping conformity, scores have been assigned as mean standardized scores to all respondents with at most two missing values. This produces valid scores for more than 99 percent of the respondents.

  • 11

    An overall factor analysis produces a single factor with an eigenvalue higher than one, explaining no less than 54 percent of the common variance. All factor loadings are higher than 0.65 and scores have been assigned as mean scores to all respondents with no more than two missing values, yielding valid scores for 96.9 percent of the respondents.

  • 12

    Response categories for those three questions are, respectively, yes/no, yes/depends/no (recoded into yes versus depends/no), and yes/no. An overall factor analysis produces a single factor with an eigenvalue higher than one that explains 40 percent of the common variance with factor loadings of 0.41, 0.73, and 0.71, respectively. Scores have been assigned to 97.6 percent of the respondents with at most one missing value.

  • 13

    Valid scores are available for 89.9 percent of the respondents.

  • 14

    Disentangling “age effects” and “cohort effects” is impossible, of course, when “one shot survey data” are analyzed. The correlation between age and birth year is then exactly −1.00. Because even with a 20-year period we face a strong negative correlation between both variables, the extremely large sample is helpful in preventing problems of multicollinearity.

  • 15

    The intra-class correlation is the proportion of the total variance that is accounted for by the contextual level.

  • 16

    We do not include additional statistical controls because a large body of research literature points out that posttraditionalism (“libertarianism,”“moral individualism,”“postmaterialism,”“cultural progressiveness,” etc.) is hardly related to other variables besides age (or better, birth cohort) and level of education on the one hand and variables pertaining to religion and spirituality on the other (see Houtman 2003 for a review of this literature).

  • 17

    A significant random effect (not included in the analyses shown in Table 4) points out that the strength of the effect of posttraditionalism on post-Christian spirituality varies significantly between the years and countries. We have tested whether this effect is significantly stronger or weaker in more traditional contexts by including a cross-level interaction between mean posttraditionalism at the context level and individual-level posttraditionalism, but this proves not to be the case.

  • 18

    The exact nature of post-Christian and posttraditionalist secularism is another seriously neglected issue in the research literature. Due to the strong focus on the decline of the Christian churches in the West, “few attempts have been made to look at the other side of the equation, at what has been called the ‘left-over,’ if one may say so” (Knoblauch 2003:268). Whereas the “decline of religion thesis” (Casanova 1994:25–35) assumes that religious worldviews are increasingly replaced by rationalist ones, there are also indications that rationalist conceptions of the self increasingly give way to postmodern ones (e.g., Gergen 1991).