SciELO - Scientific Electronic Library Online

vol.2 special edition author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Novos Estudos - CEBRAP

Print version ISSN 0101-3300

Novos estud. - CEBRAP vol.2 São Paulo  2006


Religion as a solvent — A lecture1


Religião como solvente — Uma aula



Antônio Flávio Pierucci

Translated by Anthony Doyle
Translation from Novos Estudos - CEBRAP, São Paulo, n.75, p.111-127, July 2006.




Contrary to Durkheim, for whom the role of religion is to reconnect the individual with the society to which he belongs, this essay argues that nowadays religion's social power lies in its capacity to dissolve old religious bonds and lineages. Taking Max Weber's work as its base, the text maintains that the universal religion of individual salvation, the religious form that tends to predominate above all others, works as a device that disconnects people from their mother-culture.

Keywords: religion; conversion; sociology of religion; Max Weber; Émile Durkheim.


Em oposição à visão de Durkheim, para quem a religião atua como religação dinamogênica do indivíduo com a sociedade a que pertence, este ensaio sustenta que hoje a força social da religião está na capacidade de dissolver antigas pertenças e linhagens religiosas estabelecidas. Com base na obra de Max Weber, argumenta-se que a religião universal de salvação individual, forma religiosa que tende a predominar sobre as demais, funciona como um dispositivo que desliga as pessoas do contexto cultural de origem.

Palavras-chave: religião; conversão; sociologia da religião; Max Weber; Émile Durkheim.



solvent  NOUN 1. Chemistry a. A substance in which
another substance is dissolved, forming a solution.
b. A substance, usually a liquid, capable of dissolving
another substance.

(American Heritage Dictionary)




It doesn't happen very often, but even the most experienced researchers can have their insights. I mean those fertile insights, the kind that inadvertently strike a rich theoretical payload or discover an entire empirical/phenomenal continent ripe for exploration, a discovery laden with discoveries. Likewise rare is the luck that this insight should actually “trigger” a period of intellectual productivity. But that is exactly what happened to me. I had one such insight, a short time ago. I truly did, and everything leads me to believe that I should proceed with the exploration and tap into the mine.   

It happened a year and a half ago in Berlin. In the early days of December 2004, towards the end of a stay sponsored by Capes-DAAD on which I gave a two-month Autumn course at the Free University of Berlin (Freie Universität Berlin – FU-Berlin). Far from home and with a lot of time to read and ponder, the flash came to me in plain German: “Religion als Lösemittel” – religion as a solvent.  

The course as a whole, which I gave in conjunction with my friend and colleague Sérgio Costa, was on Brazilian cultural themes and problems, and my part was focused on religious diversity in Brazil today. The title I chose for it was “Religious Diversity 'Brazilian-style'”. 

It was freezing in Berlin, and the days, bathed in only white light, were far too short. Night fell without warning at four in the afternoon. The result: a prolonged nocturnal period, endless nighttime hours without TV, perfect for reading lots and doing a little writing. I even had time to re-read some old stuff – a veritable luxury. I read a great deal on religions in Brazil, but, above all, I re-read. I re-read much of the Brazilian bibliography on the sociology of religion, a lot of things that had providentially found their way into my luggage, and other works I was luckily able to find at the Iberoamerikanisches Institut (IAI). Not much theory, but a lot of empirical research, which was good, in hindsight, as it meant I could replenish my stock of factual information on the so-called religious field in Brazil during that brief stay in Berlin, thus reinforcing my base repertoire and background ahead of that sudden flash. Which came in the last week of the course, at the beginning of December. Maybe it came to me because, unbeknownst to myself, during all that re-reading I had ended up paying much more selective attention to certain quantitative data and other empirical findings that they suddenly became more relevant and incisive than I had hitherto imagined, re-presenting themselves before my analytical will charged with “signalling signs”, shall we say, veritable traffic signs pointing me in a direction in which “religious development”2 in Brazil had been wandering down through the 20th Century but which I had never paid much attention to before, despite the fact that it already had all the projection of a very well-defined  historical tendency, empirically speaking, just as it now appeared before my eyes. 



Lately, I have been very concerned about the demographic crisis that has assailed Umbanda since the closing decades of the 20th Century. Such was the slump in numbers within Umbanda between 1980 and 2000 that the demographic health of Afro-Brazilian religions as a whole was affected by it. In 1980, the contingent of self-declared Afro-Brazilians in the census – already miniscule – accounted for a mere 0.6% of the resident Brazilian population. This slipped to 0.4% in 1991 and continued to fall up to 2000, when it reached 0.3%. “From 1980 to 1991, when you do the math, the Afro-Brazilian religions lost 30 thousand declared followers, a loss that rose to 71 thousand in the following decade3”.  In the words of a specialist on the subject: “the segment of Afro-Brazilian religions is in decline”. I would add, as they say in Minas Gerais, that “things were worse in the past, but just kept worsening”.   

Scanning the religion tables in the 2002 Census and comparing the numbers for Umbanda with those for Candomblé, only slightly more optimistic, I must confess stirred a certain – let's say – patriotic sadness. I had not expected it, I had no idea this was happening to my “Brazilian Brazil”! That Catholicism had slumped and would continue to slump was only to be expected, after all, it had been happening for some time. That Lutheranism would recede was also logical. But shrinkage within Umbanda, aka “the Brazilian religion4”, this Kulturreligion blended from indian, black and European sources, this consummate personification of our so highly-sung constitutive miscegenation, which the most patriotic social scientists, in unison with countless other patriotic intellectuals, considered the religious form most well-adapted – dare we say syncretic - to the Brazilian reality, even the most urbanized and modernized of “Brazilian realities”, now that was a truly thought-provoking, phenomenal phenomenon. 

Since the publication of the advanced tabulation of the 2000 Census data on religion, I have not been able to get this embrittling of Umbanda out of my mind, as it seems to me to be yet another depressing component of the imagetic projection of this cultural atmosphere of “bye-bye Brazil5”. Crestfallen with the dwindling figures of Umbanda in the Census, and watching our “Brazilian Brazil” also slipping toward the drain in this respect, I got down to some serious thinking: the more religious diversity broadens its scope, free from the chains of a confessional State, and multiplies the objective possibilities for free choice that a fuller stock of religions offers the Brazilian, the more pressing becomes the need for the sociologist of religion to ask himself the question:  “which kind of religion” comes off best in the predatory “cultural selection” that has taken root here – before the naked eye – since the final decades of the 20th Century? There really is no other way: in the sociology of religion one always has to start by making distinctions, or – which amounts to the same thing – by classifying. And so, as Marcel Mauss would say, let's classify (Classons donc6).  

But what classification? So many are the modalities, given that we have been stockpiling the logical tools of our discipline since Classical times, that it is impossible to know which classification to use. How does one make the right choice of sound classificatory instrument? And do so without forgetting (as if I could) that the last demographic Census of the 20th Century had brought the stark and sudden realization that between 1980 and 2000 the growth rate of Afro-Brazilian religions was negative in Brazil?7

Post-traditional ruptures with religious tradition ... post-colonial ruptures with the colonial mode of religious syncretism... so why not bring rupture with tradition back to the centre of analysis, if, as the 2000 Census clearly shows, in Brazil today, three of the most important religions sociology classifies as traditional  - Catholicism, Lutheranism and Umbanda - , each traditional in its own way, are emitting clear signs of a lack of steam when it comes to self-proliferation. This lack of stamina means that the sociological concept of detraditionalization, emphasized today by an influential theoretical vein in contemporary sociology, gains more empirical weight with each passing year, even in our paid journals, as one of the most accurate macro-descriptive concepts applicable to this religious transition – a veritable cultural tradition in fast-forward – whose developmental curve has been matched point by point by data collated by the Census Board since 1940.  

The apparently inexorable growth in conversion to Pentecostal and Neopentecostal churches of a Protestant persuasion clearly demonstrates that in Brazil today (as in Latin America from top to bottom) living a religion very often implies breaking with one's very own religious past. With this proliferation of ruptures with religious worlds that once seemed fulfilling, but suddenly no longer do, the farewells are many. One such adieu is to the Umbandist syncretism that was supposed to have fit the Brazilian cultural identity to homologous perfection.

It was because of this that I ended up discovering that it might be worthwhile to broach anew, indeed raise from the deep, that old and reliable functionalist method of classifying religions by function, which we shall call, without further ado, functional classification.  

Since my earliest forays into field work in the sociology of religion, when I started work at the old Cebrap on Rua Bahia under the grand master Candido Procopio Ferreira de Camargo, a declared functionalist amid all manner of Marxists, I learned that in order to explain what he called the “selective outbreak of religious development in the country” it is always good to cast a glance at the functional aspects of the different religiosities. One classification that he considered “very useful” in analytically tackling a field of diverse religious forces with varying degrees of developmental potential is to distinguish between the ethnic religions (which in functionalist jargon means religions whose function is to preserve ethnic subcultures) and universal religions (open to everyone, regardless of tribe, ethnicity or nationality). As Procopio wrote, in his typically concise style:     

The functional classification of religions is useful to explain the differential growth of religious forms. In this sense, we can distinguish the following: 1) those that preserve a certain ethno-cultural heritage, favouring the self-identification of a given social group; 2) those of a universal character, open to the conversion of one and all8

If there is one empirical truth the sociologist of religion is forbidden to ignore it is that there are religions and religions, and that classification and comparison are therefore of crucial importance to our discipline. One classification it is beneficial to return to is precisely this: there are religions that preserve a particular ethnic heritage and there are religions of a universal character that are “open to the conversion of one and all”, as Candido Procopio says, in words that immediately associate two ideas: openness and conversion. A “Conversion Religion” is therefore a classification that fits more readily into the second functional group, and it is thus that I return to it here, insofar as it places, at the centre of the recent boom of sociological interest in “religious modernity”, the figure – the mobile figure! as Hervieu-Léger9 would add – not only of the converted, but also of the “convertible”. By convertible we mean all those who are now and always invited to convert, that person (in the strong, anthropological sense of the word) for whom switching religion is an act of individualization, making him an individual ipso facto, a human being abstracted from inherited ties, rendered incompatible with a past that was probably no great shakes anyway, and so this individual moves – as an individual – within a religious field that is not only more plural – et pour cause –, but which is bent toward plurality, bombarded as it is by a plethora of regular and unregulated religious choices.   

There is another aspect in this that is best expressed in the more conventional jargon of sociology: what we see in religious conversion is the shift from an ascribed (religious) status to an acquired (religious) status. Conversion, the switch from the religion into which one was born to a religion of one's choosing, is first and foremost an act of social mobility. If we go a little further in our theoretical treatment of this openness to mobility that ends up affecting the various religious contenders – though perhaps not all of them - I believe we can reformulate Candido Procopio's definition in the more ambitious terms of a thesis designed to firmly identify the sense of this “openness to all”:    

A religion is considered universal when its structure is geared towards converting the person into an individual. (Proposition 1)

With fresh emphasis, one can see how heuristically useful the functional classification defended by Candido Procopio continues to be, shedding a most opportune light upon the current dynamics of the field of religion, “at least in Brazil”, as the old master modestly intended to show10. Yet it is by no means exclusive to Brazil, as we well know.

Besides the heuristic validity inherent to this functional classification, which has fallen into disuse, the book Catholics, Protestants, Spiritists, written by Candido Procopio and his team at Cebrap11, contains a table in which the “Brazilian religious institutions” are distributed between these two functions. This table in itself is enough to ensure the continued relevance of the work. Looking over it now, thirty-something years later, the least one notices is that certain religions have changed function since the list was compiled, stepping out of their roles as preservers of an “ethno-cultural heritage” to adopt a “universal character”, as we shall see.  

Religions of a universal character: The Apostolic Roman Catholic Church; the Brazilian Apostolic Church; the Anglican; the Baptist; Congregational; Episcopal; Lutheran; Methodist; Presbyterian; Independent Presbyterian and Pentecostal Churches; Assembly of God; Christian Congregation of Brazil; Church of the Quadrangular Evangel; The “Brazil for Christ” Pentecostal Evangelical Church; The Seventh Day Adventist Church; The Adventist Church of the Promise; The Mormon Church; Kardecist Spiritism; Umbanda; Theosophy.    

Religions with the function of preserving ethno-cultural heritage: Judaism; the Armenian, Greek, Russian and Syrian Orthodox Churches; Islam; Buddhism; syncretic sects linked to the Japanese colony and its descendents (Seicho-no-Yê, World Messianic Church, etc); Candomblé; Xangô; Indigenous Religions12

What has changed in this panorama since 1973? Apart from the brute and conspicuous absence of the various Pentecostal denominations that have since emerged, principally the Neo-Pentecostal churches (the Universal Church of the Kingdom of God, the International Church of Divine Grace, etc.), one striking aspect is that Candomblé and Umbanda feature separately. In 1973, Umbanda ranked among the “religions of a universal character”, while Candomblé and Xangô were listed among those for the “preservation of ethno-cultural heritage”. An interesting conclusion: when applied to Afro-Brazilian religions in 1973, the functional schema served to separate Umbanda and Candomblé, placing them in different categories of social function.     

Not today. Today, Afro-Brazilian cults as a whole, even the most Africanised strains of Candomblé you sometimes stumble upon during fieldwork, behave in large measure, albeit not entirely, as universal religions, no longer restricted to the ethnic or racial markets13.



The anthropologist Livio Sansone recently launched a book entitled Negritude without Ethnicity (2004). What we are now seeing is something more or less akin to that – a kind of “Afro-Brazilian religiosity void of ethnicity” currently creeping into Candomblé and associated traditions, such as Batuque, Xangô and Tambor de Mina. 

As we know, since its creation in the 1920's, Umbanda has claimed to belong to all Brazilians: black, indian or white. It has been that way since the cradle, or rather, since its conception, as the intention was never for it to be a black religion or a religion exclusively for blacks, quite the contrary, according to its motto: the Umbanda is for all of us14. On the other hand, Candomblé, lived and transmitted as “a religion for the preservation of specifically black cultural heritage”, as Prandi15 points out, gradually veered towards “a religion for all”, consequently spreading throughout large Brazilian cities during the 20th Century, including many in the southeast, where, prior to 1960, it was little but a regional reference for immigrants thinking back up north to Bahia16.   

One point deserves particular attention: the shifts detected by the functional schema follow a pattern. A clearly “developmental” pattern, shall we say, thus avoiding the pitfalls of the term “evolutional”. The changes in function that have been occurring within the Brazilian religious field all reveal a shift in the same direction: from ethnic religion to universal. This reverts the trend witnessed with the universal religions brought to Brazil by European and Asian immigrants, the most illustrative cases being the Lutheranism of the German colonies and the Buddhism implanted by the immigrant Japanese. Over the course of the 20th Century the pattern that emerged was of religions shedding their function of preserving ethnic identities in order to become universalist in their salvational missions and universal in their scope of recruitment. I repeat: no longer the contrary.   

This change of function and consequently of target population, which Candido Procopio and his team had already identified in the dynamic exhibited by Lutheranism17 “in Brazil”, eventually stumbled upon the ethnic stronghold par excellence in Candomblé and its sister cults. To say nothing of the “new” faiths connected with the Japanese colony and its descendents, such as Seicho-no-Yê, etc., which ended up undergoing the same process of ethnic unblocking



With the problem framed in such verifiable and precise quantitative terms, scientific curiosity requires that we investigate the colour composition of these Afro-Brazilian religious groups, which are vulgarly still considered 'black religions' even today. Well, this need for a reality-check one day transformed into a decision to quantify the colour composition of religious creeds in Brazil and, vice versa, also the religious composition of each colour contingent identified by self-declaration. Reginaldo Prandi and I did just this, using a data sample of nationwide scope kindly supplied by the statistical company Datafolha, which has been collecting such data in presidential election surveys ever since18.       

I will concentrate here on the data referring to Afro-Brazilian religions, as this is my focus and I believe it will be sufficient to provide an empirical base for my arguments.

As common sense would dictate, blacks are indeed the greatest relative presence in Afro-Brazilian religions. While they comprise only 7.3% of the total Brazilian population, they account for 18.0% of the Afro congregation. The percentage of mulattos is even higher, at 29.2%, which means that blacks and mulattos combined make up 42% of the so-called “saint folk”. Almost half, then, but no more than that. However significant their slice of the pie chart may be, it is still not a majority. At 51.2%, it is a simple irony of life that the absolute majority in Afro-Brazilian religions is white!    

Taking Umbanda separately, the white majority is even more expressive, reaching 56.6% and therefore pushing the black and mulatto proportion down to 42.7% (15.2% and 27.5% respectively). In Candomblé, however, the balance tilts in relation to Umbanda, with blacks and mulattos comprising 56.8% (23.9% and 32.9% respectively). At 40%, whites account for a minority, but an obviously numerous one, large enough to deconstruct once and for all the old image of Candomblé as a religious stronghold of the blacks. Candomblé continues to be a collective bearer of an African past that is paradoxically no longer the real past of the majority of its adepts when taken individually, no matter how Africanised they may be in religious terms.    

It's obvious, right? Ça va sans dire that the whites of Candomblé are not there to celebrate their primordial roots in an African ethnicity. Much less to “defend”, for some absurd reason, their racial and social identity as whites, which would be pure bravado, to say the very least. Their African identity is, in truth, an Africanisation through faith, what Weber calls a “purely religious” identity, which is why it is able to co-exist with the real ethnic identities of 40% of the followers of Candomblé.   

To hone in on the problem of the growing disjunction between religion and ethnicity and to be better able to contemplate the paradox of an ethnic religiosity whose roots a large part of its congregation can never share, it will be interesting to view the flipside of the same data. Having looked at the colour breakdown of each religious creed, we will now examine how religious affiliations are distributed within each declared colour. 

Herein lies a detail that is crucial to my argument: taking total conversions to evangelical churches, we can see a slightly larger percentage of mulattos than whites (14% against 12.6%), but a far higher percentage of blacks (16.8%). In other words, statistics for the turn of the Century (20th into 21st) in Brazil show that the percentage of converts to evangelical churches in general and to the Pentecostal churches in particular is far higher among blacks than among mulattos and whites, in that order. Otherwise put: if, on the one hand, the number of black Evangelicals in Brazil today is still not larger than that of black Catholics, far from it indeed, the fact remains that there are much more black Evangelicals than black Kardecists, black Umbandists, blacks in Candomblé or blacks in the “other religion” or “no religion” categories. If we were to delve still further we would find that the number of black Pentecostals (14.2%) is significantly higher than white (9.1%) and Asian19 Pentecostals (6.3%), and even of mulatto Pentecostals (11%).

Leaving the Evangelical milieu to one side, let us look at the demographic presence of blacks in Afro-Brazilian religions? In a predominantly Catholic country, where 75% of whites declare themselves Catholics and 76% of mulattos, the percentage of black Catholics is curiously lower, standing at 70.1%. If we do the math - 70.1% of blacks are Catholic, 16.8% evangelical, 2.7% Kardecist spiritist, 1.7% adherents of other religions and 5.7% stating “no religion” – that leaves a meagre 3% of the black population – that's 3% and not a tad more! -  who declare themselves to be adepts of the Orishas. According to the Datafolha survey, this 3% breaks down into 1.7% Umbandists and 1.3% practitioners of Candomblé.   

In other words: when we initially focused on the colour distribution within each religion, we found exactly what we expected, that is, the religion with the highest participation of blacks is indeed Candomblé (23.9% of adherents declared themselves black). However, when we looked at the religious distribution within each colour grouping, it took just a little extra attention to see that the number of black converts to the Pentecostal churches (14.2%) far outweighed the number of stated adepts of the religions of the Orishas (3.0%).      

And, as there is a side to our curiosity that is only ever satisfied by pure numbers, a brief return to the tables of the 2000 Census is enough to raise one further numeric shock: in the year 2000 there were 1,675,680 black evangelicals, that's nearly one million, seven hundred thousand – close to two million black evangelicals, against a mere one hundred thousand (95,521) black adepts of the Afro-Brazilian cults (66,398 in Umbanda and 29,123 in Candomblé). That's a big difference!    



A universal religion is basically defined by its “openness to all”. But this basic definition does not fully address the problem. The mere idea of openness to all does not strike me as sufficient. It, alone, does not seem capable of capturing the disruptive implications – which are what interest me – of an actively universalist religiosity, exemplified by the monotheist religions with their universal missions, religions whose constitutive dynamism is founded upon belief in a mission issued by Divine revelation and which is effectively none other than to pursue the universal propagation of the revelation itself. The mission is to proselytise, to preach, opportunely or inopportunely, to not rest until the conversion (or submission, according to Islam) of all whom God has called has been achieved.      

Naturally, for this kind of religion, any stress placed on a shared ethnic identity, any trace of commitment to a particular people or population, any civic-particularistic cultural attachment becomes an uncomfortable impediment at the very minimum, or downright unthinkable at most. Localism makes no sense when it comes to universalising Divine Grace (or Devine glory, at least). That's why radical Protestantism, with its resolutely individualist appeal, lends itself much less to any ethnic attachment than the likewise radical monotheism of Islam. While Islam would like to see the whole world submitted to Allah, the one and only God, it also harbours recurrent exclusivist tensions, which would like to have the Arabs recognised as a particular people, like the Jewish in Judaism20

It is also clear that, in Brazil, of all the religions with a universal vision and mission - which do not cease to increase in number and variety -, it is those of an individualist and militant missionary bent that prove most successful. In other words, in a structure of cultural fields that are becoming increasingly more competitive both within and amongst themselves, the Protestant congregations have a clear edge. Sociological theory, from the most classical to the most contemporary, is there to provide us with the key to understanding this well-known and highly special comparative advantage that Protestantism enjoys over its rivals: Protestantism is the religion of individual conversion par excellence, and radically so. In Protestantism, voluntary adherence is much more highly valued than a religious heritage or “lineage of faith”; what really counts is being “born again” into a flock of “the spiritually reborn”. The meaning is in personally joining a group of true believers, and the most radical formulation of this principle resides in the puritan concept of the Believers' Church21.       

At this point it seems to me that we are close to finding a good explicative key: producing individuals by dissociations is a sine qua non of a purportedly universal religion's unimpeded diffusion. A religion has to be individualizing in order win converts outside the prophet's own group, as, after all, no man is a prophet in his own land. 



Irreplaceable given its relevance today, we cannot disregard, much less underrate, Weber's methodological/individualist contribution to the sociology of religion. With each passing day his schema seems to me even more current, mainly when I see myself faced with the new challenge of tackling Brazilian religious modernity theoretically in the midst of this global hypermodernity. In short, in an essay whose theme is religion as a solvent, there could be nothing better than to earmark in Weber's work exactly those moments in which he sticks his finger in the (incurable) wound of the (irresolvable) tension between the universal religion of individual conversion and the domestic or family sphere. Weber's sharpest formulation of this irreconcilable tension is in the essay “Intermediary Consideration”, dedicated to the irresolvable conflicts between the final values, immeasurable in and by all, that govern the legality of each cultural sphere; an essay whose subtitle - “Religious Rejections of the World and Their Directions” - is rather better known than the title itself. A quote:  

When salvation prophecy created communities on a purely religious basis, the first power with which it entered into conflict was the community in its naturally given form, the clan, which it feared could undermine it. He who cannot antagonize the members of his household, his father and mother, cannot be a disciple of Jesus: “I do not bring peace, but the sword” (Mt 10:34) is uttered in this context (and, mind, in this context alone). Of course the vast majority of religions has also regulated intra-worldly ties of filial piety. But the wider and more internalized the scope of the concept of salvation became, the more clearly it emerged that the believer should be closer, above all, to the saviour, to the prophet, the priest, the confessor and the brethren than to his kin and the matrimonial community in their pure sense. With this at least relative devaluing of these relations and rupture with the magical ties and exclusiveness of kinship, the prophecy, especially when transformed into congregational soteriological religion, has developed a new social community. [eine neue soziale Gemeinschaft].22 

A magnificent, extraordinary (if not insolent) Weberian moment, with all the sheer sociological cruelty of “a heartless specialist”. Speaking of religion dissolving the bonds of family, the bloodties that we all hold in some way sacred, another version of this brilliant sociological masterstroke appears in the chapter “Sociology of Religion” in Economy and Society23.  

One way or another, what I want to show with this veritable Weberian bulls-eye, which I hand-picked especially, is the background argument that unveils, in all its capacity to disturb, the implacable individualizing logic inherent to the concept Weber's sociology of religion develops of the congregational salvation religion. In the interests of brevity I will restrict my commentary to the last line of the passage, which speaks of “congregational soteriological religion” as the creator of a “new social community”. What I want to say, with a little lifting, is that it looks like what the great classic German thinker had in mind was not the “elementary forms of religious life”, as had Durkheim, but the “elementary forms of new religious life24”. Approaching the religious phenomenon from a standpoint diametrically opposed to that of Durkheim, Max Weber held a privileged view of his subject not as something that consolidates the past, the inherited and the ascribed, sanctifying them from and for the inside, but as something that enflames them from the inside, open as it is to the outside and to the future, to the invention of a new community life, sought, tested and chosen by the “now-individuated-individual” who has allowed himself to respond to a calling, an invitation, an enunciation – may I even say a summons – in which he sees himself and converts. 

What we have here is a particularly disruptive if not effectively destructive form of religion. Indeed, disruptive, predatory and extractivist. At the outset, a religion of individual salvation can only swell its ranks through extraction; by systematically plucking members from other congregations, of which, prior to heeding the “good news” calling them to apostasy, they had felt structurally and restfully part and parcel. But no, conversion religion has no consideration whatsoever for such feelings. It accentuates the part and undoes the 'we', it shatters inherited social relations and dismembers already established collectivities. Congregationalist in kind, it herds together (“con-gregare”) individuals that it disaggregates from other flocks, whether by succession or abduction; individuals it recruits by uprooting them, de-territorializing them from their conventional settlements, leading them from their conventional paths, systematically undermining other religious belief systems and ways of life, criticizing and condemning lifestyles and behavioural codes, whether religious or not, collective or not, significant or not, without the slightest ceremony.     

Universal religion of individual salvation frees people from the established community routines and disentangles them from the ready-made weft of communication and subordination just so, once individualized, that is, liberated and autonomized, it can engage them as individuals in the construction of a new community, in fieri, which has only purely religious bonds to offer them, vertical and horizontal religious ties whose pure religious specificity means they have to appear precisely as they are, endowments that bestow an entirely distinct, new and  altogether other subjective meaning: one that is, as Weber says,     ausschliesslich religiös25.   I see passages like this as important peak moments in Weber's sociology, moments in which he gives wings to his hardly consoling vision of how certain types of religio-cultural configuration foreshadowed the dynamic of modernity26.  

Heedless of the past and geared toward the future, the very encapsulation of religious modernity in nuce, congregational religiosity is the elementary form of religious life; a fact of which Max Weber was convinced and which I, today, have the satisfaction of re-presenting with fresh emphasis.  



The object of sociological interest that Weber set forth in his Religionssoziologie is very different to that established by Durkheim. What really interested the latter was identifying the religious dynamic at the heart of those new fraternities of followers that, when taken to the extreme, bubble over into sects. In other words, his focus was on sectarian “communalization”.  The word 'sect' itself, it is worth remembering, comes from the Latin secare, meaning “cut, or separate by cutting” 27.    

With no prior formation through any extra-religious ties, congregational religions come together from zero as specifically religious associations of individuals poached from other groups (religious or otherwise). One can see why the adverbs exclusively and purely are so strategic to Weber's characterization of what he calls religiöse Vergemeinschaftung (a religious “communalization”) as opposed to religiöse Gemeinschaft (religious “community”). In effect, if what we are dealing with here is the ab ovo formation of a religious community, that is, of a community lashed together and kept together by specifically religious bonds, then the first step necessarily has to be this: dissolve all other ties that bind these individuals to their previous communities, whatever they may have been; cut them loose from their former community shackles, whether religious or otherwise.  What Weber makes us see is that, from the outset, there is a tug-of-war between the “new bonds” (purely religious fraternity based on “the brotherhood of faith”) and the “old bonds” (those of fraternity based on kinship, colour, locality or patria, or on an inherited, native and now depreciated religion). The ties that must be severed, and that includes those of a symbolic nature, are those that belong to the past, thus leading to an iconoclastic approach; conversion, after all, means changing religion, and this sometimes requires a certain dose of iconoclasm.    

The image of the sword in the words of Jesus cited by Weber could not be more appropriate when the order is precisely to cut ties as apparently solid and sacred as those of the blood. Weber views this as a war cry, a call to cultural war, that is, that pits the sacred against the sacred28. Razing the inherited “is basic”, in the modern sense of the phrase. So unlike the totemic religion a la Durkheim, we find ourselves before the “congregational” religion of Weber, which works, first and foremost, as we can see, as a solvent. This is because – and here I can even call upon Parsons in my defence, when he sums up as follows his reading of Max Weber's sociology of religion– “Weber's main interest is in religion as a source of the dynamics of social change, not religion as a buffer of social stability29”. As “monkey” Simão likes to say in his column in the Folha de S. Paulo newspaper – “More direct would be impossible”.

What universal religion targets and wants is the individual, and so it produces individuals, it turns them out. The individual stripped of all ties is precisely the simple unit they wish to collect (re-ligare, or 'bind again'), something that is only done using the 2nd person singular. What they seek is not the person as a pulsating cortege of ready-made relations, a complex tangle of social positions with all their conventional obligations, a paid-up member of an integral, holistic collective unity; what they want is not the person, but the individual as a single unit. With this I hope to make clear, once and for all, that when Weber speaks of religion as “congregational religiosity”, he means “of individual conversion”.    

This highly precious Weberian notion (so sociological, though entirely antipodal to Durkheim's vision) – for showing itself to be such an accurate prognosis of the future developments that would fuel this form of religion's growing predominance over the course of modernity, so realist avant da lettre in wisely seeing through to how corrosive a modern religion can be, so self-evident in its heuristic validity that its thunder clap rouses little or no attention or intellectual interest  - cannot, in my view, continue to be ignored by the Brazilian social scientists who dedicate themselves, through scientific vocation, to the scientifically sociological study of our nation's religions. Given its in-born radicalness, I consider this idea both strategic and indispensable if we are to get a good angle on the religious template that nowadays permeates the most important eruptions of confessional congregational growth – modern and post-modern – in the religious hotbeds of Brazil and all other nations of Latin America.   

The religion that interests the Weberian sociologist of religion is not that which, prior to all else, celebrates a pre-established community. Quite the contrary, his core religious element of interest is precisely the religion that creates, that has the charisma (whether personal or institutional) to invent brand new communities ex nihilo, even if that means severing prior ties and consequently throwing the continuity of the old, standing community geography into disarray. As such, though not explicitly stated, Weber clearly pits his sociology of religion against that of Durkheim.   

As religion, for Durkheim, “is above all a system of notions through which individuals represent themselves before the society of which they are members30”. Durkheim spells it out clearly here, leaving no room for doubt: “la société dont ils sont membres”.

It is enough to go back to Weber's systematic study of “sociology of religion” (the title of Chapter V of Part I of Economy and Society) to see that the subtitle perfectly and literally brings out the formal object of what ought to be a sociology specifically of religion. English does not have words nearly as processual or synthetic as religiöse Vergemeinschaftung (religious “communalization”) to convey, sociologically, the construction process of new religious communities with purely religious ends and drives. Unlike for Durkheim, the relevance for Weber does not lie in associating the idea of religion with pre-established groups equipped with a repertoire of integral, rapturous and extra-quotidian rituals with which to celebrate the supra-individual and sanctifying force of a pre-ordained vital cohesion that, being social, can only ever be generically religious and never, as Weber envisages, driven by the logical demands of a specifically religious conceptual scientificity.   

Individual religious conversion as an individualizing rupture with once consolidated bonds is a total inversion of the form of religion that had such an impact on Durkheim's later sociology, namely a religion that uses ritual to sanctify and thereby psychologically replenish the cohesiveness of an old and given human grouping.      

I know that in order to better explore this contrast between these two classic systems of thought I ought to rummage more thoroughly through the conceptual universe of Durkheim (beyond, I must stress, his mental universe, which is excessively taken with the socially cohesive dynamism of the tribal or national religion), but the limited time available to me does not permit such a foray. For the time being, then, I must be content to highlight once again that the contrast not only exists, but is, especially on this point, extremely stark. But first I would like to call upon the support of two highly respected social scientists who attack Durkheim's oversocialized (remember Dennis Wrong) conception of religion from different flanks. 

The first and earliest criticism was lodged by Raymond Aron (1967) in this somewhat  terse passage from his compendium Main Currents in Sociological Thought

Durkheim says that society creates religion when in a phase of effervescence […]Individuals find themselves in such a psychic state that they feel drawn by impersonal forces that are at once immanent and transcendent […]accounting for religions in this way summons a causal explanation based on the premise that social effervescence is favourable to the emergence of religion. [...]Now, as soon as the cult turns toward society, it can assume no other form than that of a tribal or national religion. As such, the essence of religion would be to inspire men to fanatical adherence to partial groups and to at once sanctify the individual adherence of each to the group and the group's hostility towards all others. Defining the essence of religion in terms of the adoration the individual devotes to the group certainly strikes me as inconceivable, especially when the adoration of a social order, at least to my eyes, would seem to constitute the very essence of impiety. To suggest that a transfigured society is the object of religious sentiment degrades rather than saves the human experience of which [the]sociology [of religion]purports to give account31

The other, extremely recent criticism is that of Clifford Geertz in a text to be presented at the colloquium Les Sciences Sociales en Mutation, scheduled for this year in Paris and organized by the Centre d'Analyse et d'Intervention Sociologiques (Cadis) in partnership with Le Monde.  At a certain point in his text, entitled “La religion, sujet d'avenir”, an unwitting Geertz mentions Durkheim with no other purpose than to show how his conception of religion is out of step with the meltdown of modern times:     

Though it has often been, historically speaking, the most deeply-rooted cultural structure in its place of origin, and the one whose expression is most clearly marked by the local conditions, religion has become something fluctuant – and increasingly so -, stripped of social anchorage in a pregnant tradition or established institutions.   In place of and instead of a solidary community aggregated by collective representations (Durkheim's dream), what has emerged is a Simmelian network, diffuse and centreless, connected by generic, multi-directional and abstract ties. Religion has not weakened as a social force; quite the contrary, it would seem to have been strengthened in recent times. But it has changed – and continues to change – in form32   

Availing of this stepping stone in which Geertz bids a summary adieu to religion a la Durkheim, I would like to modestly add that, at least in Brazil today, one of the most salient aspects of this “social force” religion continues to exert (not just any old religion, however, but one very specific type) lies precisely in its statistically proven capacity to dissolve old ties and sever the established religious lineages. Bye-bye Brazil!   



After giving so much thought to the boisterous numbers concerning religious conversion collated by the Brazilian census and other quantitative surveys, it suddenly occurred to me – see how the world turns – to take another look at Althusser's famous thesis on the functioning of ideology in his celebrated essay of 1970 “Ideology and the Ideological State Apparatus”, namely that “Ideology interpellates individuals as subjects”.  

Unexpectedly inviting, Althusser's theses on ideology had the immediate effect of giving me the urge to develop a similar formulation to describe the totally un-Durkheimian modus operandi of what we could call the elementary form of religious modernity, in other words, the foreshadowing, even if only in seminal form, a thousand years in advance, of the spirit of modernity in religion. All I had to do was shift the position of the word “individuals” from the point of departure of Althusser's thesis to the point of arrival of my own to achieve the following echoed formulation:

“Universal religion interpellates people as individuals” (Proposition 2)

Basically, like ideology in general, it needs to address the individual (“Hey, you!”), not the group as a whole, much less the group member. Quite the contrary, in fact, because the latter is its raw material in the production of individuals, i.e.: members who leave the religions into which they were born in order to become parts of another religion, one in which the value of the bonds is not measured in inheritance, lineage or ancestry and whose promise of salvation is not to redeem the collective from plague, war or drought, but to liberate the individual, as an individual, from the status quo of the group.  

The universal religion of individual salvation, the religious form that tends to predominate over all others in the general development of culture, functions as an instrument of Vergemeinchaftung, of the community in fieri, which must, if it is to form, disconnect people from their mother cultures, from what would seem to be their natural and therefore congenial cultural context. It plucks the indian from his tribe and the neighbour from his neighbourhood, turning the stranger into one's true bosom buddy, one's true brother in the bonds of faith, protracted and projected into another relationship with the temporality of social ties, that bundles one's kith and kin into the past and projects upon the stranger the fellowship of the future.   

Having rummaged through Weber in search of a theoretical key to developing my Berlin insight, I ended up finding it in another classic German thinker, Georg Simmel; another key to dissolving – note that I have eschewed the word 'resolving' – the Durkheimian enigma of religion as the dynamogenic reconnection of the individual with the society to which he belongs. This key, even if it does not unlock all the doors, certainly opens a large window onto the idea of religious modernity. It is basically this: for Simmel, the connection religion represents and re-presents is not with society, but with the species – an horizon so thoroughly universal that the particularistic notions of a society “of which we are members” or of “our group” find themselves entirely out of their depth. Let us finish on a note from Simmel:      

We need to consider the feeling of dependence that has come to be seen as the essence of all religion. The individual feels connected with some universal, with something superior, whence he came and to which he shall return, to which he differs but is nonetheless identical. All of these emotions, which converge upon a focal point that is the notion of God, stem from the individual's relationship with the species33.   

For Simmel, the connection is with the human species; it is with humanity, not with the tribe-society or nation-society “of which we are members”, as Durkheim had thought.

So you want a universal cultural solvent? Try a soteriological congregational religion, prescribes Weber. “Try, and pay up”, add our Neo-Pentecostals.  



[1]A lecture delivered on April 20, 2006 as part of the selection process for a lecturing post in the Sociology of Religion at the Humanities Faculty of the University o f São Paulo. 

[2] In Weber's terminology, religiöse Entwicklung, cf. WEBER, Max. “Die Wirtschaftsethik der Weltreligionen: 'Einleitung'”. In: WEBER, M. Gesalmmelte Aufsätze zur Religionssoziologie, I (GARS I). Tübingen: Mohr, 1988 [1920], p. 245.

[3] Prandi, Reginaldo. “O Brasil com axé: candomblé e umbanda no mercado religioso”. Estudos Avançados, nº 52, vol. 18, Sept-Dec. 2004, p. 226.

[4] Umbanda, uma religião brasileira (São Paulo: CER-FFLCH/USP, 1987 [col. Religião e Sociedade Brasileira, vol. 4]) is the title of a renowned work by the São Paulo sociologist Maria Helena Villas Boas Concone.

[5] Pierucci, Flávio. “'Bye, bye, Brasil': O declínio das religiões tradicionais no Censo 2000”. Estudos Avançados, nº 52, vol. 18, Sept.-Dec. 2004, pp. 17-28.

[6] Mauss, Marcel. Sociologie et anthropologie. Paris: PUF, 1950, p. 335 [Braz. transl. Sociologia e antropologia. São Paulo: Cosac & Naify, 2003, p. 371.]2003: 371.

[7] Most intriguing of all is that the data on religious affiliations recorded in the demographic census are collected through self-declaration, which means that the loss of adepts within Umbanda is a loss of declared adepts, despite the climate of politico-cultural religious freedom that grew over the same twenty-year period. One would assume that such politico-religious liberty should have been a good thing for the different religions, not the opposite. Nonetheless, a decreasing number of Brazilians nowadays declare themselves religiously Afro-Brazilian before the Census Bureau.  

[8] Camargo, Candido Procopio F. de et al. Católicos, protestantes, espíritas. Petrópolis: Vozes, 1973, p. 23, my highlights.

[9] Hervieu-Léger, Danièle. Le pélerin et le converti: La religion en mouvement. Paris: Flammarion, 1999.

[10] We can therefore dispense with Candido Procopio's disclaimer, offered out of excessive caution and theoretical modesty, that “the concrete application of this criterion of distinction is only meaningful in the Brazilian socio-cultural context”. (Camargo, C. P. F., op. cit). It is curious, however, that in expressing this doubt, he had in mind the case of Buddhism “in Brazil”, a universal religion that was still an “immigrant religion” for us in the 1960s; that is, a closed, private religion restricted to the Japanese immigrant population and their first descendants. “Buddhism in Brazil constituted an expressive example of how a large universal religion selects functions in order to preserve the ethno-cultural heritage of the Japanese immigrants and their descendants”. (Idem, p. 23).

[11] Idem.

[12] Idem, p. 22.

[13] Cf. Prandi, R. “Raça e religião”. In: Prandi, R. Herdeiras do axé: Sociologia das religiões afro-brasileiras. São Paulo: Hucitec, 1996.

[14] The Umbanda is for all of us is the title of a Master's thesis defended in 1967 at the University of Wisconsin by the demographist and sociologist Maria Stella Ferreira Levy.

[15] Prandi, R., op. cit., p. 64.

[16] “Yoruban or Nagô Candomblé (along with the Ketu, Ifa and, later, Pernambucan Nagô ritual variations) and Angolan Candomblé will come to establish themselves in São Paulo, not as religions for the preservation of black cultural heritage, nor as ethnic religions, but as universal religions; that is, as creeds open to all, regardless of colour, origin or social background”. Prandi, R. Os candomblés de São Paulo. São Paulo, Hucitec, 1991, p. 20, my highlights.

[17] During the compilation of this two-column list, his mentor aptly pointed out that, in Brazil, Lutheranism was a religion that had once fulfilled the role of “preserving the ethno-cultural patrimony of the German contingent and their descendants” (Camargo et al, 1973). By the early 70s, Lutheranism “in Brazil” had already joined the ranks of the universal religions.   

[18] The scope of the Datafolha survey was restricted to the Brazilian electorate, not the entire population. Even so, as voting is mandatory and the electoral register is highly valued in Brazilian political culture, almost 90% of Brazilians of voting age feature in the Electoral College. For this reason (cf. Berquó, Elza & Alencastro, Luiz Felipe de. “A emergência do voto negro”. Novos Estudos Cebrap, nº 33, June 1992, pp. 77-88) the research sample we used comes very close to representing the country's entire youth and adult populations.  

[19] The Orientals are the ethnic group with the lowest number of evangelicals (8.3% are Pentecostal and 2.0% Protestant). In compensation, after Catholicism, the second largest religious category among Orientals is “other religions” (10%).

[20] Cf. Bruce, Steve. “Religion, Ethnicity, and Social Change”. In: Bruce, Steve. Religion in the Modern World: From Cathedrals to Cults. Oxford/New York: Oxford University Press, 1996, pp. 96-128.

[21] Cf. Weber, Max. A ética protestante e o “espírito” do capitalismo. Transl. José Marcos Mariani de Macedo. São Paulo: Companhia das Letras, 2004, pp. 230-231 (Edited by Antônio Flávio Pierucci).

[22] Weber, Max. “Zwischenbetrachtung [Consideração intermediária]: Theorie der Stufen und Richtungen religiöser Weltablehnung”. In: Weber, M. Gesammelte Aufsätze zur Religionssoziologie, I (GARS I). Tübingen: Mohr, 1988 [1920], pp. 536-573. [Braz. transl.: “Rejeições religiosas do mundo e suas direções”. In: Weber, M. Max Weber: Textos selecionados. São Paulo: Abril Cultural, 1980, pp. 237-268 (Os pensadores Collection)].

[23] Weber, Max. “Religionssoziologie”. In: Weber, M. Wirtschaft und Gesellschaft. Tübingen: Mohr, 1985 [1922], p. 350. [Braz. transl.: “Sociologia da religião”. In: Weber, M. Economia e sociedade. vol. I. Brasília: Editora da UnB, 1991, p. 386.]

[24] The adjective I have inserted into this classic Durkheimian title is shamelessly lifted from the title of a book by Roy Wallis, which, precisely in the study evoked here, sought to disentangle the issue of conversion to the New Religious Movements by examining it in the light of a new category created by he and Steve Bruce: the “structural availability” of the converted (cf. Wallis, R. Elementary Forms of the New Religious Life. Londres: Routledge & Kegan Paul, 1984).

[25] Translation: exclusively religious (Weber, M. “Religionssoziologie”, p. 276).

[26] Turner, Bryan S. Max Weber: From History to Modernity. London & New York: Routledge, 1992, p. 34.

[27] Cf. the glossary in the new edition of The Protestant Ethic and the “Spirit” of Capitalism, where we find the following: “Two etymological origins are attributed to the word 'sect': from the Latin verbs sequi, meaning 'follow, come/go after, obey' and secare, meaning 'cut, separate by cutting, divide'. Through a highly exclusionary process, the sect separates and poaches individuals from their natural communities, social networks and prior value-sets and immerses them into a new group context which demands the total adherence of the individual and control over others”  (2004, pp. 289-290).

[28] What could be more sacred than human blood? Isn't that how Durkheim also presents it? 

[29] Parsons, Talcott. “Introduction to Max Weber”. In: Weber, Max. The Sociology of Religion. London: Methuen, 1965, p. xxx.

[30] Durkheim, E.. Les formes élémentaires de la vie religieuse. Paris, Quadrige/PUF, 1998.

[31] Aron, Raymond. Les Etapes de la pensée sociologique. Paris: Gallimard, 1967, p. 361.

[32] Geertz, Clifford. “La Religion, sujet d'avenir”. Le Monde, 04/05/06.

[33] Simmel, Georg. “A Contribution to the Sociology of Religion”. In: Simmel, G. Essays on Religion. New Haven & London: Yale University Press, 1997, p. 115, my highlights.



Received for publication on June 2, 2006.



Flávio Pierucci is a lecturer in sociology at the University of São Paulo.