The ‘secularization thesis’ states that religion declines as societies modernize.

The characteristics of modernization are widely agreed upon, though
different thinkers have emphasized different aspects: Max Weber emphasized
rationalization, Karl Marx emphasized industrialization, and Émile
Durkheim emphasized social differentiation. All these founders of modern
social science agreed, however, that religion and modernization were
inversely related. By positing a necessary link between modernization and
secularization, the secularization thesis defines religion as a relic of the
past, and explains continuing religious practice by reference to an incomplete
or unaccomplished modernization. Religion is conceptualized as something
that humankind must be liberated from, or as something that ‘holds
back’ such things as progress and science.

Yet much empirical evidence
actually points the other way: many of the most rapidly modernizing
societies – i.e., those across the ‘global south’ – are among the world’s
most religious societies.2 In the United States, meanwhile, modernization
seems to be closely correlated with religious vibrancy.3 Typically, data of
this kind has been accounted for by saying the Western European modernity
provides the template for modernization; other parts of the world will
eventually, it is held, converge on the European model. It increasingly
seems to be the case, however, that Western Europe is not the model but
the exception. If Western Europe is indeed a largely secularized society,
this may be because of the specifics of its own history, not because it
provides a universal template.4 In some cases, perhaps, modernization is
good for religion. But if this is the case, then secularization must either be
uncoupled from modernization or redefined as something other than
religious decline.

In his book Public Religions in the Modern World, José Casanova
distinguishes among three definitions of secularization: secularization as
religious decline, secularization as privatization, and secularization as
differentiation. Of these, Casanova concludes, only secularization as
differentiation is a defensible hypothesis.

By ‘differentiation’ Casanova means
the fact that autonomous secular institutions now handle functions that were
formerly under the domain of the church. Economy, law, and medicine are all
obvious examples. Casanova insists, however, that differentiation does not
necessarily entail decline or privatization. Indeed, he concludes on the
basis of empirical evidence that religion worldwide is not in decline, and
that we are witnessing a widespread rejection of the privatization component.
For him, the Iranian revolution of 1979 is a watershed moment.

Etymologically, one of the meanings of secularization referred to the
transfer of individuals and property from the church to the mundane
world. The focus of secularization so understood is on a changed relationship
within the world, rather than on a changed relationship between
this world and the next. This has immediate ramifications for a concept
such as Abrams’s ‘natural supernaturalism’, which aims to describe
precisely a changed relationship between the natural and supernatural
worlds. It is clear that Abrams was thinking of secularization primarily
along Weberian lines, presumably because Weber’s emphasis on rationalization
provides an easy link to an interpretation of romanticism as a response
to enlightenment reason.

Following Weber, then, Abrams’s notion of
‘natural supernaturalism’ (derived appropriately enough from Thomas
Carlyle, the great nineteenth-century opponent of rationalization5) implicitly
defines religion as a set of ideas and presumes that modernization entails
the privatization and/or loss of those ideas. Thinking of secularization as
differentiation, by contrast, would mean that in principle one has nothing
to say about personal belief; differentiation is concerned simply with
describing how religion adapts itself to the accelerating emancipation
of secular institutions and norms.6 We can imagine, therefore, a highly
differentiated society that nevertheless experiences high levels of religious
participation.

What of the ‘privatization’ of religious belief? The demand that
religious expressions and subjectivities be limited to the private sphere can
take several forms. In one narrative, privatization is the natural result of
differentiation: there is simply less and less in our public life that needs to
be described in religious language. Another version of privatization,
however, involves the deliberate application of state power, for example in
the command that church and state be separate, and that religion not
‘invade’ the public sphere. This is a policy known in France as laïcité. As
Olivier Roy points out, laïcité developed ‘against the backdrop of a
political conflict between the state and the Catholic Church that resulted
in a law regulating very strictly the presence of religion in the public
sphere’ (xii). It was to laïcité that the French appealed when they banned
conspicuous religious symbols from the public schools in 2004 (a law
widely interpreted as aimed specifically at Muslims). Roy argues that one
source of present-day French confusion about religious minorities is that
laïcité in effect treats Islam as the new Catholicism.

The Invention of ‘Religion’

It seems logical to assume that the word ‘religion’ points to some thing in
the world. We commonly take religion to be a stable object – a ‘natural
kind’, as cognitive scientists would say. Yet consensus is growing that
‘religion’ is not in fact a natural kind, but rather that it is a historically
constructed, or invented, object. The evidence comes from several quarters.
A number of cognitive scientists have argued that religion is not in fact
selected for by evolution but is rather a ‘by-product’ of other evolutionary
pressures.7 More immediately relevant to literary study, historians like
Peter Harrison and Jonathan Sheehan have demonstrated that ‘religion’ as
we understand it today is a historically malleable category, forged in the
early modern period to answer particular historical and conceptual
needs. As for religious studies, a number of scholars have demonstrated
that the discipline, which depends centrally upon such concepts as ‘world
religion’, has constructed ‘religion’ as a transhistorical, universal category
available for study and analysis.8

Peter Harrison traces the emergence of a modern understanding of
religion to the various crises of authority within Western Christianity
during the early modern period. Harrison notes first that in the aftermath
of the Protestant Reformation, it became increasingly important for
believers to grasp precisely what it was they were professing to believe.

Moreover, the content of that belief was a particular kind of knowledge,
a packet of information leading to salvation:

‘the traditional view had been
that in the process of revelation God reveals himself. Now God reveals
saving knowledge’ (25).

Second, Harrison shows that understanding the
content of ‘religion’ as a set of propositional truths, and ‘belief ’ as assent
to those truths, made it possible to handle the fact of religious pluralism
in a new way. Alternatives could now be understood as different religions,
in the plural, to be ranged against the ‘true religion’ (that is, Christianity).
Thus ‘religion’ was placed on an equal footing with the natural sciences,
and ‘comparative religion’ became for the first time conceivable.
There are many details to fill in here, but perhaps I have said enough
to indicate the main points.

There is substantial scholarly agreement that:
(1) ‘religion’ is not a natural kind; (2) It was invented to answer particular
needs at a particular historical moment; (3) This history centres on a crisis
of authority in Western Christianity; and (4) Inventing ‘religion’ also
makes it possible to invent ‘religions’, in the plural, in order to name those
activities and postures that characterized Europe’s Others.

Secularism

Just as we often speak of ‘religion’ as a generic category, so we often speak
of ‘secularism’ as the opposite of ‘religion’. But if religion is not in fact a
natural kind, what of secularism? Is it too invented, and can we narrate
its history? To ask these questions is to invoke what is at present an
extraordinarily wide-ranging, interdisciplinary conversation.

In an influential essay entitled ‘Modes of Secularism’ (1998), Charles
Taylor suggests that the seventeenth-century wars of religion, culminating
in the peace of Westphalia in 1648, are the ‘origin of modern Western
secularism’ (32). By establishing the modern system of European nation
states, the Peace of the Westphalia recast religious dispute as a matter of
internal state politics rather than of inter-state conflict. The secular, which
in an earlier era had largely been conceptualized temporally – as the
mundane time that would come to an end with Christ’s second coming,
and as the time presided over by secular institutions – was now conceptualized
spatially. Once there are secular spaces, it becomes possible to
draw conceptual boundaries around them, and to conceive of religion as
something that might violate those boundaries.

Taylor goes on to clarify two distinct historical modes of modern
secularism. The first, derived from Locke, Leibniz, and the deist tradition,
seeks to identify ‘common ground’ among the world’s religions. (Note
that this development goes hand-in-hand with the invention of the
category of ‘religion’ described above.)

The second mode, associated with
Grotius, develops an idea of secularism as an independent political ethic
abstracted from religious beliefs. Taylor astutely points out that much of
the confusion surrounding religion in the public sphere today has its
origin in these different conceptions of the secular.

In the first, secularism
has its origins in religion; it seeks a lowest common denominator that can
bring subjects into agreement, and the state’s role is to be evenhanded
among a variety of religions.

In the second, secularism has its origins in
a non-religious theory of the human; it seeks to secure a space free from
religion, and conceives of the state’s role as the active policing of religion
(Taylor 35).

Finally, Taylor notes that because both models were developed to
manage disputes among disagreeing Christians, neither can readily handle the
variety of metaphysical orientations on offer in the contemporary world.

As a result, Taylor proposes a third mode of secularism, this one modeled
on John Rawls’ notion of an ‘overlapping consensus’. An overlapping
consensus model of secularism acknowledges the fact that all people have
a substantive conception of the good, but it lifts the requirement that
those conceptions be the same, or even compatible. Taylor writes:

‘It is essential to the
overlapping consensus that it be generally understood that
there is more than one set of valid reasons for signing on to it’ (49).

Thus,
X might support a functionally secular state because she is an atheist and
believes that government should not be in the religion business, while Y
might support a functionally secular state because he doesn’t want the state
interfering in his religious life. Meanwhile, Z doesn’t care much about
religion one way or the other but as a transgender activist is committed
to human diversity, and sees in a secular state the best chance to protect
and promote that diversity. And so on.

Such a re-conceptualization is
necessary, according to Taylor, because modernity has brought with it
a new ‘social imaginary’. We no longer live in a hierarchical, heavily
mediated society; rather, we imagine ourselves as having ‘direct access’ to
power in the context of a homogenous secular time shared by all. This
shared secularity is where any theory of secularism must begin.

In his 2003 book Formations of the Secular, Talal Asad offers an
important criticism of Taylor’s overlapping consensus model. Asad is doubtful
that modern democracies are in fact ‘direct access’ societies. While access may
not be mediated by explicit hierarchies, he notes, it is nevertheless
mediated by all manner of other things: pressure groups, mass media, elite
leaders and administrators, and so on.9 These forms of mediation are
certainly different from those that obtained during the era of Latin Christendom,
but Asad argues that we mis-describe that difference if we see it
as simply a shift from a religious to a non-religious imaginary. Rather, the
post-Westphalian secular state has in a variety of ways taken over functions
formerly intrinsic to religious authority.

Now, there is widespread agreement that Christianity (particularly
Protestantism) in some way ‘causes’ secularism. Asad’s point, however, is
that we should not interpret secularism as progress. The secular state
authorizes certain subjectivities and disallows others, legislates practice and
bodily posture, and pursues its own interests with regard to those citizens
it identifies as ‘religious’. Influenced by Michel Foucault’s concept of
‘governmentality’, Asad calls attention to statecraft as the science of managing
a population in which power is always operational but often invisible
because it works locally, in those places where individual bodies and practices
– wearing a headscarf, eating halal meat – intersect with institutions.

When it comes to secularism, then, the key term for Asad is neither
‘consent’ nor ‘reason’ (as found in liberal political theory) but rather
power – particularly power as enacted and mediated within the modern
nation state. If we bring these two together, colonial and post-colonial
history immediately becomes relevant to the discussion. Just as ‘religion’
was invented partly in response Europe’s others, so too was secularism
invented, legitimized, and naturalized through an encounter with its other
– an ‘other’ most recently figured as the Islamic fundamentalist.
We can establish four general points of agreement among those who
aim at a minimum to flush secularism from its hiding place of impartiality
and neutrality:

1. Secularization needs to be carefully defined and cannot be understood
as simply a subtraction story, as if the modern secular self was always
there, waiting to be liberated from superstition. This premise is often
linked to broader claims about ‘multiple’ or ‘alternative’ modernities.10
2. Secularism is not a neutral governance structure but has its own interests.
It authorizes certain kinds of subjects and marginalizes others. It is
coercive (but frequently disguises this under the name of ‘tolerance’).
So to analyze secularism we need also to analyze power.
3. ‘The Religious’ is not the opposite of ‘The Secular’. Rather, secularism
is complexly intertwined with a particular religion (Christianity) and as
part of that complex relationship produces, at a certain historical
moment, the distinction between the religious and the secular. As a
result, religion appears as marked, set against the neutral or unmarked
background of the secular.
4. Secularism is a product of a particular historical process in the West. It
does not travel very well, and it is unlikely that it can be plunked down
somewhere else. Turkey and India are frequently cited examples of the
difficulties of this process.11 Asad makes the point somewhat more
polemically when he argues that state secularism guarantees peace at
home ‘by shifting the violence of religious wars into the violence of
national and colonial wars’ (7). However one chooses to trace the
connections, it seems clear that the question of secularism is simultaneously
a Western question and a global one.

To be sure, there are important disagreements even among those who
agree upon these basic premises.12 Take, for example, the relationship
between secularism and Western Christendom. In A Secular Age (2007),
Charles Taylor offers an immensely detailed and rich account of this
relationship; Taylor’s argument is that Christianity in effect created secularism
over roughly that past 500 years. Yet he offers his account without
much attention to the other cultures and other ‘religions’ that were
increasingly pressing in upon Europe’s consciousness during this same
historical period. And elsewhere in his work he remains at least cautiously
optimistic that the western model of secularism, problematic as it may be,
is the best hope for other regions of the world.

At the other extreme, Gil
Anidjar has recently taken the same premise – namely that Christianity
created secularism – to argue that secularism is therefore the new name
for a centuries-long western persecution of Islam, now understood as the
archetypal ‘religion’: ‘Secularism is part of a discourse of power and of
institutions that are bent on making us . . . know or recognize religion . . . for
what it is not: Christianity, secularized’ (62). There are a range of possible
positions between the poles laid out by Taylor and Anidjar.13

Despite their manifest differences, it bears repeating that Taylor and
Anidjar share the same basic (and critical) premises regarding secularism.
There are many others, of course, who wish to defend secularism – and
the liberal tradition with which it is importantly linked. Thus Jurgen
Habermas, though he has moved recently toward a fuller engagement
with the continuing salience of religion in the West, continues to insist that
religious adherents ‘translate’ their values into language that is recognizable
to secular actors. Within romantic studies, Mark Canuel’s groundbreaking
Religion, Toleration, and British Writing (2002) is virtually alone in recognizing
the institutional dimensions of secularism, and thus deftly sidestepping
matters of belief and intention that have hampered most writing on the
topic. Canuel’s position on romanticism’s discourse of toleration, and his
faith in institutional mechanisms to solve or at least manage religious
dispute, puts him in the broadly Habermasian camp.