Limits of Religious Analogy: The Example of Celebrity

Focusing on celebrities is often compared to a religious behavior, be it by the actors when describing their own practices or by scholars when using analogies with “cult”, “sacralization” or “sanctification”. Such comparisons appear to be both obvious and hardly convincing, since they merely evoke, without analyzing or explaining. Moreover, they ignore the normative effects—be they positive or negative—produced by any kind of religious analogy. This paper proposes several paths toward a reasoned use of comparison with religion: extending comparison to differences and not only to resemblances; passing from “religion” in general to the plurality of religions; deconstructing the said “religious” phenomenon into several functions depending on contexts; replacing discontinuous categories by continuous typologies and, finally, “religion” conceived as an original matrix by “religion” conceived as a contextual configuration. “Religion” thus appears as a common sense notion rather than as a conceptual instrument, and analysis may then fully develop without being restricted by religious analogies, while comparison may be used as a real tool.


Introduction
Religious analogies have been flourishing since the beginning of the academic literature on celebrity. 1 In his pioneering book on stars, French sociologist Edgar Morin often proposed "equivalences"-to use his own term-with religion, among which the supposed homology between stars and sacredness ( [2], p. 77): "So begins the at once mystical, radiant and smiling ascent of the 1 For a fully developed version of this paper, see [1].

OPEN ACCESS
stairs. Equivalent to the Roman triumph or to ascension of the Virgin, this ceremony comes back every day", he writes about the "red carpet ceremony" in Cannes ( [2], p. 98). As for Violette Morin, she preferred Greek polytheism to Christianity, choosing to baptize "Olympians" the modern targets of a "personality cult" [3].
What can we do with such analogies, which abound both in common sense (ordinary people and journalists) and among scientists (sociologists, anthropologists, historians)? Do they demonstrate real affinities between traditional religious behaviors and admiration for famous people, so characteristic of the modern world? Or are they just metaphors, actually more evocative than descriptive or analytical? What do they bring to, or what do they take out of, the knowledge and understanding of this phenomenon? Should we see it as a "substitute for religion", according to an already overused formula, that is to say a "cult" and not just an element of "culture"? And, more generally, under what conditions are religious comparisons epistemologically justified? We will try to answer these questions, first, by presenting some examples of analogies between celebrities and religious figures, and their often blurring or critical effects, even when they seem to be helpful; then, by distinguishing between two meanings of "comparison", and by turning the common sense term "religion" into a systematic analysis of the various "functions" it encompasses, through the example of fame; and finally, by proposing to consider "religion" not anymore as an original matrix, but as a mere contextual configuration.

Scholarly Comparisons: Misery of Analogy
One generation after Morin's essay, many books dedicated to celebrity keep on using "religious" words in their titles, as noted by French sociologist Philippe Le Guern, who himself published on "fan culture": "Many explicit titles (The Adoring Audience, The Saints and the stars, Elvis is alive, Resurrection of the King, The Presley cult, Media cults, Cult Television…) suggest a contamination of the field by a religious terminology" ( [4], pp. [25][26]. Presley in particular arouses a strong drive toward analogy among sociologists and anthropologists: he seems to be the object par excellence of religious comparison, as noticed by John Frow [5] or Gabriel Segré ([6], p. 3). In an article significantly entitled "Is Elvis a God?", the former presents "half-gods like Diana or Elvis" as "intercessionary figures, gods in human form whose presence spans and translates between two worlds" ( [5], p. 202). Nonetheless, he notices that "these analysis tend to partake of the very religious ethos that they describe" ( [5], p. 201), thus highlighting the ambiguity of the researcher's position when carrying out such assimilations: "The religious relation exists midway between dead metaphor and a theory which has yet to find itself "( [5], p. 200), all the more since "in none of these analyses (and the countless others like them) is there a clear sense of how seriously the concept of the sacred should be taken, and of how the sacred might have come to be caught up with" ( [5], p. 201). One is tempted to follow him when he criticizes those analogies in that "the notion of the charismatic genius and of the atavistic reincarnation each in its own way inscribes itself within an imaginary of ineffable presence" ( [5], p. 201).
Suffice to say that one never exactly knows where the researcher is in proposing these analogies with religion: does he share them with the actors, in an empathic or mimetic relationship (but then, what does such an analysis bring that we would not already know?) Or does he produce them by an interpretation that the actors would not spontaneously possess (but they are the first to use a religious vocabulary, albeit at a lower level of generality?) And if these analogies are relevant, then how can we account for the fact, however obvious, that these figures which are called "religious" do not officially belong to the religious corpus as recognized either by religious authorities or by the faithful? (an anthropologist who investigated on Claude Francois's fans in France aptly remarked that he "is not explicitly considered a saint, even by his most fervent admirers" ( [7], p. 296). Should we assume that what is at stake is a religion in the making, or a resurgence of an ancient religion? But in this case, what kind of religion: paganism, animism, christianism? And if it does not resemble any of the listed religions, then what evidence is there that we are dealing with a religious phenomenon?
Even if these analogies can produce a certain effect of enlightenment, if not of revelation, one wonders to what extent they shed light on the very experience of the actors, or on the expectancies of the researchers who observe them. This issue is all the more worth addressing that it pertains, more generally, to the methodological and epistemological status of interpretation in the humanities and social sciences.

A Blurred "Religion": The All-Aspiring Effect
"This ethnographic issue of the Festival as a presentification test of the film world would perhaps possess a certain kind of affinity with a certain way to address the issue of religious acts, or better, of the religious institution as an organizational form of the divine presence", cautiously suggests a French historian ( [8], p. 47). However, why would it not rather be "of the political institution as an organizational form of the royal presence"? It is strange indeed to refer this ceremony exclusively to the religious sphere, whereas it can pertain as well to the political one. Moreover, if it is true that Cannes festival deals with "issues of presence, testimony of faith, strength of adhesion, consensual participation of community members, etc.", should we really use "the language of liturgy or theology" ( [8], p. 47), notwithstanding that these behaviors can be met as well in political rallies or sport competitions? As French historian Paul Veyne noted, "any fervor is not religious": "Stalin's 'personality cult' did not have the vibrato of transcendence but of charisma, and it would be exaggerated to say that Marxism or communism were a kind of religion since they do not bear but one drop of such thing. I have never seen any votive offerings to Stalin in order to find a lost object, no more than to Elvis Presley" ( [9], p. 89). By the way, equivalents of votive offerings can be observed about celebrities; but it is nonetheless necessary to take into account their differences with recognized ex-voto, named and treated as such [10].
Religious analogy seems to be used by some scholars as an exclusive interpretative tool, transforming the research object into a mere projection of their cognitive categories. Everything then tends to become "religious", by a kind of all-aspiring effect that attracts into "religious" anything that more or less looks like it, without ever discussing the relevance of such an assimilation, without any definition of what it means. Simultaneously, "religion" tends to be reduced to what has been chosen as the focus of observation: for example, "recognition and belonging", which allows British sociologist Chris Rojek to wonder whether "the celebrity culture has replaced religion" (but if religion is reduced to recognition and belonging, then not much in human experience might remain outside "religion"); or else, "the rites of ascent and decline" that organize celebrity culture, as a result of "primitive religion" ( [11], p. 416)-but can they not also be found, among other domains, in the army? Therefore, everything becomes a proof of the modern idols "holiness", including the theft of their stele (as in the case of James Dean, Dylan Thomas, Sylvia Plath, Buddy Holly and Jim Morrison ( [11], p. 494)-even though the theft of saints' headstones is not exactly a documented practice in the history of Christianity.
Similarly, this inflation of the space allotted to "religion" leads many specialists of "celebrity culture" to speak of "belief", whereas the simple question "belief in what?" should have questioned the relevance of such a term. Thus, in spite of the remarkable qualities of his research on Elvis Presley's fans, Gabriel Segré cannot resist to mention a "self-construction as believer" ( [6], p. 177): but believer in what? What is at stake here, obviously, is worship; but does it rely on a belief, and if so, which one? If the actors have the freedom not to precise the exact content of their beliefs, the researcher is supposed to define the terms he uses.
Still more strikingly, American sociologist Michael Jindra, in his study on Star Trek fans, states that the phenomenon is "profoundly religious" even if it lacks a belief in a deity or a supernatural being, and compares it to the Church of Scientology in that it also has a fan club, a dogma, a recruitment system, and relies on ceremonies [12]; but he forgets to remark that none of these characteristics is specifically religious.

Forgetting the Critical Function
Now let us listen to someone who fully belongs to the religious world: French bishop Jean-Michel di Falco. "Television, he said, has its 'saints', the stars to whom the audience turns in order to solve a problem. And when leaders or great media figures are thrown out, one speaks of 'excommunication'! (...) Being on TV means to enter the holy of holies, a place of supreme legitimacy, which is equivalent to a consecration. I deliberately use this religious vocabulary". Indeed, television is so "impure" today, because too popular, that referring to religion here is not meant to understand the phenomenon but, rather, to stigmatize it. In the scholarly world to which the bishop belongs, describing a profane practice in religious terms always has a critical dimension [13].
Far from being new, this phenomenon has been highlighted by historians of theology and of Hellenism: an idol is a "bad simulacrum", referring to an "archaic" thought, if not explicitly "pagan" ( [14], p. 36). Idolatry is somehow the religion of the others, that is to say, of pagans: "The initial dichotomy between Christian and pagan idolaters has been replaced by a new distinction that vaguely opposes ourselves and the others. And an idol remains either the image of others, or another image" ( [14], p. 194). Drawing down onto religion, whatever its possible relevance, obviously have a demarcating function, drawing a line between "us" and "others", between true Christians and quasi-pagans, or between laymen and scholars, illiterate and literate.

The Embarrassment of Analogy
Few researchers criticized these religious analogies used by specialists of celebrity culture: their approximations, their lack of reflexivity, the way they blur the very definition of "religion" even if it is supposed to provide a key to the phenomenon, their tendency to swell the religious sphere, their spontaneous syncretism that makes them mix up quite heterogeneous religious traditions. However, French sociologist of media Eric Maigret is clearly critical about it [15]. He notes the implicitly normative use of the word "cult": either as a stigmatization of the "popular", spontaneously equated with "barbaric practices, socially or psychologically regressive"; or, conversely, as a valuation of their "initiatic" and "transgressive" dimension, "based on communion and celebration" ( [16], p. 103).
But can we simply dismiss these analogies for their lack of scientificity? For nevertheless, it is difficult to do without: they come to our minds, they seem to make sense, to tell something, and they sometimes give the feeling that something right has been stated, that some aspect of the phenomenon has been enlightened. Moreover, as we share a one and same culture with the actors we study, the use of empathy has some legitimacy: our intuition that there is something relevant in these analogies cannot be totally unfounded, since we are as familiar with this world as the people we observe-a short: we are really embarrassed. "Religion is embarrassing for us, embarrassing for me", admitted Frow ([5], p. 207).
How can we get out of this embarrassment? Certainly not by simply euphemizing terms: as Rojek prudently stating that "celebrity culture is no substitute for religion. Rather, it is the milieu in which religious recognition and belonging are now enacted" ( [11], p. 97); as Matt Hills, a British specialist of "fan" culture, preferring to see in the "fan" attitude a form of "neo-religious", where assimilation to religion is produced by the actors in order to give more consistency to a seemingly irrational behavior [17]; or even as Gabriel Segré, who speaks of "recomposition of the religious" about the homologies between Elvis's "cult" and properly religious cults [6]. Although laudable, these attempts to circumvent the pitfalls of analogy explain neither where the pitfall is, nor what it precisely helps to understand in the phenomenon.

A Survey of Experimental Psychology
Can empirical investigation help us? In an experimental study devoted to "the relationship between worship celebrities and religious orientation", a team of British psychologists, assuming that there is a "worshiper personality", tried to compare the scores of a test on attitudes towards celebrities with religious dispositions.
The results are not spectacularly successful: they show that "while religiosity grows for men as for women, the tendency to 'worship' celebrities decreases", but also that "some individuals have high scores on one and the other scale" [18]. Interestingly, the authors do not remark that both correlations, positive (worshiping celebrities growing with religious dispositions) and negative (both decreasing together) may indicate an affinity between these two attitudes: in the first case, if worship is supposed to exist in unlimited quantities (in which case there may be high scores on both scales), and in the second case, in limited quantity (in which case one goes down while the other goes up). The only valid counter test of a similarity between these two attitudes would be a lack of correlation, be it positive or negative.
Should we deduce that this loose two-way correlation advocates for analogy? The authors do not seem to be quite convinced. They rather stress the importance of responses suggesting that these conducts are nothing more than a mere "entertainment": fans would be "primarily attracted to their favorite celebrities because of their ability to entertain" ( [18], p. 1168). However, to consider this survey reliable, we should be sure that the test measures something else than the capacity of respondents to satisfy the implicit values contained in the questions: namely, the low degree of legitimacy of the celebrities "cult", which prompts them to minimize their level of commitment. Even an uneducated fan knows that he will look less foolish when declaring that he listens to Elvis's music for pleasure rather than when declaring that he prays him in order to have his vows fulfilled.
As for the interpretation proposed by the authors of the negative correlation between religiosity and worship celebrities, in terms of obedience to the Christian precept not to worship any god, it also shows a certain naivety. It postulates, first, that religious dispositions would imply to strictly follow clerical requirements; and second, that actors consciously assimilate celebrities to gods, as the researchers do as a hypothesis: two quite risky postulates. In short, if the study does not, as the authors conclude, confirm the hypothesis of the existence of a "worshiper personality", it does not confirm anymore the opposite. That method is not good.
It is thus time to raise the level of our reflection by clarifying, first, the status of comparison, then that of "religion".

Compare "to" or Compare "with"? The Inventory of Differences
If comparison is, according to Durkheim, the laboratory of social sciences, yet we need to be clear as to what it exactly consists in. This requires distinguishing between two meanings: the lay one and the scholarly one. French historian Paul Veyne [9] highlighted this distinction, emphasizing the difference between, on the one hand, "comparing to" (that is to say, assimilating a thing to another, because of their similarities) and, on the other hand, "comparing with" (that is to say, making explicit the similarities and differences between two entities). Only the latter sense-heuristic and not ontological-is relevant for the work of historians, sociologists, anthropologists etc., since it avoids the excesses of uncontrolled analogies, literary metaphors, valorizing or, conversely, stigmatizing assimilations.
This distinction is important: for example, comparing Stalinist crimes to Nazi crimes produces a political delegitimization of communism, whereas comparing one with each other is the legitimate work of historians. However, it is not always well accepted, not only by the actors-who spontaneously engage in comparing to-but also by some researchers. Here, any comparison between a phenomenon pertaining to celebrity and a religious figure should necessarily bring out the similarities and differences between these two areas.
Of course, one can compare, as Jake Halpern does ( [19], p. 160), the enthronement procedures of a star in the Hollywood Walk of Fame to the secret negotiations by which Catholic Church decides a beatification or sanctification [20]. However, something crucial would be lost in such a comparison: namely, the fact that the concerned institutions are not religious but civil; and that they pretend to honor someone who has been distinguished not by his/her virtues, sufferings or faith, but by his/her fame, talent or beauty-which is not an insignificant difference. Similarly, it is tempting to consider the managing company of Presley "cult" as a "kind" of church; but if what is at stake is a "religious" phenomenon, then why those who are concerned do not just go into churches, pray saints if they are catholic or create new sects if they are protestants?
Of course, this question is not sufficient in itself to challenge the assumption of a religious phenomenon: after all, some beings not listed as saints or deities may properly be treated as such precisely because they escape such an official status [13]; but at least, it should engage an analysis of this fundamental difference. Is it not that we miss an essential dimension of the phenomenon if we forget-due to the blindness produced by the projection of such an analogy-what it would need to be truly religious?
A possible answer to this question is that the "cult" of celebrity would be a weakened form of religiosity, which fails to reach the status of a full religion fault to possess its qualities, worth, elevation. Already before, in the history of Christianity, Church authorities used to separate the "bad" popular devotion-iconolatrous and idolatrous-from the "good" faith in its authorized forms [21]. The same in modern times with scholarly attacks against mass culture, especially with Frankfurt School when interpreting the differences between high and low culture in terms of lack or loss [10].
In short, cultural ethnocentrism explicitly denounces the "cult" of celebrity as a form of discount culture and, implicitly, as a misguided or debased form of religiosity. Conversely, the opposite clan valorizes these popular practices, eventually using religious analogy in order to offer them the cultural legitimacy they do not possess. In both cases, nothing has been described, analyzed nor understood: there has only been a reduction to religion or an enhancement by religion. It is thus time to get out of normativity in order to finally get into the real research job.

Compare with: to Finish with "Religion"
Highlighting differences as well as similarities is therefore the first key to a really heuristic, rather than normative, comparison. The second key, as a consequence, is to drop the general category of "religion" in favor of a detailed analysis of the components of the phenomenon thus designated. Indeed, one cannot highlight the differences and similarities without decomposing the observed item into a plurality of constituent features, more or less specific.
A first step is to move from "religion" (or "the religious") to "religions", thus taking seriously the differences between religious traditions. Uncontrolled comparisons between celebrity and religion tend to mix up, sometimes in the same sentence, references to Catholicism, polytheism, Buddhism, animism. Such syncretism would make sense if we were able to subsume these categories under a common concept of "religion", to which we could provide a clear and consensual meaning, broad enough to encompass all its possible forms and, at the same time, precise enough not to include phenomena that are not specifically religious. However, such a definition is obviously missing in the toolbox of the social sciences, despite a few scattered attempts that apparently convinced but their authors. Let us quote here Peter Berger's definition of religion as "the establishment, through human activity, of an all-embracing sacred order, that is, of a sacred cosmos that will be capable of maintaining itself in the ever-present face of chaos" ( [22], p. 51): but who will define the "sacred", and who will explain how this definition distinguishes "religion" from "culture"-since the latter may well, too, be regarded as a form of sacredness?
Rather than regretting the absence of such definition and adding to the many attempts to fill the gap, we'd better postulate that there are good reasons for such an absence: the main one being that, as Max Weber observed, the content of an abstract category has all chances to emerge inductively, following the work of description and analysis, rather than by a definition a priori [23]. This brings us back to the need for a fine decomposition of the components of the phenomenon designated by the actors as "religious". Owing to this operation of desubstantialization, it appears then not as an essence but as a contextual "amalgam" of more or less specific properties, as suggested by Veyne: "Almost all the components of the amalgam could exist outside of a religion and be expressed differently" ([9], p. 486).
This, of course, is not entirely new: Durkheim had proposed a similar program one century ago ( [24], p. 49). In his perspective, the religious phenomenon can be reduced neither to the supernatural and mysterious, nor even to the idea of god: the relevant criteria are the "dual division of things into sacred and profane" ( [24], p. 50); the existence of these two basic types of phenomena that are beliefs, as forms of representation, and rituals, as forms of action; and, to support these beliefs and rituals, the institution of a Church, which differentiates between religion and magic.

A Functional Decomposition of Fame
In the path thus opened, we can go on one step further and deconstruct the phenomenon called "religious" not anymore by criteria, as does Durkheim, but by functions. Function has to be defined here according to relational and efficient dimensions, as that by which an entity does something to another entity. Thus it becomes possible to detach such or such a function (e.g., the separation function) from the structure (e.g., religion) in which it is contextually implemented, and conversely. Then the researcher's analysis may become considerably more opened and softened: this or that function may have been implemented in this or that religion but not in another, or in non-religious contexts; this or that religion may include this or that function, but not another. Exit therefore "the religious", and welcome to what might be called "bouquets of functions", the composition of which varies according to contexts.
Such functional decomposition somehow resembles French sociologist Danièle Hervieu-Léger's analysis when she distinguishes four religious dimensions: ethical, cultural, communitarian, emotional [25]. The latter function is particularly relevant for the topic of celebrity, since it gives a place to charisma, a property common to traditional and modern societies, characterized by a "charismatic thrust" in relation to the media: "The central place of charisma in the recomposition of the religious field is, generally speaking, a hallmark of societies strongly engaged into modernity" [26]; and "this charismatic thrust in the religious field of advanced societies has to do with the phenomenon of personalization and mediatisation of powers, partly due to the power of media" ( [26], p. 346). Here we turn back to the issue of celebrity not anymore through a comparison to religion, but by highlighting similarities between distinct functions: cultural, communitarian, emotional or charismatic.
French sociologist Albert Piette proposes a similar approach about what he calls "secular religiosities", defined by three main traits: representation of a transcendent (supernatural) reality, able to deal with death; specific "ritual" set; and "sacralisation" of some persons, ideas or objects [27][28][29]. This third feature is at the heart of the behaviors associated with celebrity, while the first two ones are only marginal (the soteriological function is not precisely central in the cult of celebrities), or partial (if admiration rituals may exist, they are quite dispersed according to the categories of stars, and little institutionalized). We see here how comparison with religion, even when not fully conclusive, may be an excellent analyzer.
French anthropologist Philippe Descola also prefers to think in terms of functional decomposition, given the profoundly unsatisfactory nature of any a priori definition of religion, even those of Durkheim, Mauss or Dumezil ( [30], p. 78). He thus highlights a particular function, common to art and religion: figuration, defined as "the public establishment of an invisible ontological quality by a statement, a prophecy, an image" ( [30], p. 82). He then distinguishes three classes of "invisible establishables": "spirits", "gods", and "predecessors", which are divided into "ancestors" and "totems" ( [30], p. 85). A central place is thus offered to "religious image", which is present "in domestic icons as well as in the Lobi ancestors altars or in the central Brazil masks" (but also, one might add, in works of art as well as in pictures of celebrities): namely the image which "makes clear the identity of the prototype and its reproduction. Figuration is the appropriate term" ( [30], pp. 86-87). We touch here a nodal point of the "cult of celebrities", which finds its origin and its dynamics in the referential game between the presence and absence of the representative and the represented [31].
Without attempting to provide an exhaustive list of functions involved in the various religions, let us only mention, first, those common to fame and to a particular religion, and second (true to our program of comparison with), those found in one but not in the other.
If we use Durkheim's functional decomposition, what may be the meaning of a "sacralizing" or "separating" function in the case of celebrity? It appears in the "apart" status of celebrities [10], but in a fairly informal way: for lack of "temples" in due form, the establishment of a collectively recognized cutoff between the world of celebrities and that of bystanders hardly materializes in ceremonies, when metal barriers or security guards block any circulation between ordinary people and their objects of admiration. Moreover this separation is not sustained by any prohibition other than that-common to all citizens-pertaining to the respect of privacy.
Two other central functions in the phenomena identified as religious are also involved in the case of celebrity, but in a minor or little formalized way. The ritual or ceremonial function only appears in collective journeys celebrating the death or birth anniversaries of the idol, or, less dramatically, with the waiting at the concerts exits or at awards ceremonies. As for the sacrificial function, it can be met only in extreme cases, when a fan is devoting all his/her resources, time and energy to pilgrimages or to buying objects related to his/her idol. Now, the charismatic function is central in celebrity. It involves the ability of "idols" to trigger phenomena of intense collective emotion-whatever its exact causes. The cultural function is also present, with shared references, stories, media communication, special behaviors (seeking autographs, accumulating information, gossip...). The communitarian function is implemented through fan clubs and gatherings for a concert or ceremony. We also noticed the importance of the figurational function (especially under its totemic form, following Descola's analysis [30], each "idol" serving as a "totem" for the small community of his/her admirers). This figurational function is of course bound up with the aesthetic function of creation or incarnation of beauty, central in the case of music or film performers: a function that appears in most religions, through chants, ritual objects, ceremonies, even if it is not specifically religious since it also appears in secular contexts-as well as, indeed, the other three functions just mentioned [10].
Much more central however is the cultual function, according to which admirers or worshipers demonstrate-through actions, words, emotions-their attachment to a collectively invested object.
Here is an obvious contiguity with hagiographic practices in the Catholic tradition. Even if celebrity "cult" is much more informal and unsupported by authorities, even if there are no actual altars erected to "idols" other than native houses and Halls of Fame or Walk of Fame, the fact remains that many individual behaviors pertain to what is called a "cult", from collecting objects evoking a celebrity to searching autographs, from repeatedly contemplating his/her image to searching his/her presence or regularly talking of him/her with those who share the same passion, or participating in pilgrimages on one's tomb [6]. In short, the need to admire or worship, which at the time of triumphant Catholicism found a moral form in the cult of saints, manifests itself today in the praise of the talent or aesthetic qualities of such or such an actor, singer, or "personality" [10].
Another property sometimes referred to by fans is the ability of their idol to be a "guide to life", a "star that shows the way", helping, supporting, giving strength, consoling. This is the ethical function, implemented by those who are invested with a moral authority (or, sometimes, by those who challenge moral values). Though much less central than in Christian religion, the moral dimension is not absent of fame, but it is specifically designated as such only about categories of people whose fame is a quality added to their status, such as political figures. However, here too, the link between religion and morality is in no way absolute ( [32], p. 542): thus this minimal moral dimension of celebrity does not mean that it possesses a religious character.
This brings us to another category of functions: those documented in Christian tradition but which do not exist, or only marginally, in celebrity culture. The latter is not related to survival in afterlife, except through the posterity offered to "great singular ones" [13]: the soteriological function of salvation is thus excluded, at least for fans (but it is also the case with paganism, at least according to Veyne, for whom salvation is not a specifically religious item ( [9], p. 531). Above all, celebrity does not require belief in supernatural properties or in any regime of invisibility as deity, spirit, soul: that excludes or at least severely limits both the function of mystical contact with the beyond, and the thaumaturgic function of miracle-although in some extreme cases (including Presley cult), a reference can be done to almost miraculous healing abilities. Finally, passion for celebrities goes very well without the institutional function of a church (although, again, some fans clubs may tend to enter into a process of institutionalization)-which is not, admittedly, a slight difference with "religion".
To complete this sketchy comparison, let us note that there is at least one characteristic associated with celebrity that does not appear in the various practices normally associated with religion: namely this form of, one might say, "negative worship" that is the culture of gossip, often demeaning or insulting, associated to multiple intrusions into the privacy of celebrities [33,34]. We are dealing here not with the lack of a "religious" characteristic, but with the presence of a non-religious characteristic-which raises quite a different definitional problem.
Finally, the most typical functions in terms of celebrity worship are the cultural, charismatic, communitarian and figurational ones: except for the first one, none is specifically religious, since they also appear in secular contexts. Less obvious, or only in extreme cases, are the sacralising, ritual, moral and sacrificial functions. Marginal or exceptional are the mystical, thaumaturgical and soteriological functions. The institutional function is almost absent.

From Original Matrix to Contextual Configuration
Our analysis cannot in any case support the conclusion that the celebrity "cult" would be a modernized form of religion, or even a return to the "elementary religion", this "tribal religion" that was, for Durkheim, totemism. Speaking of "recomposition of the religious" or "drift of the sacred" or "secular religion" [27], or of "disenchantment of the world" (or its opposite, as John Frow suggests [5]), means using, once again, the paradigm of "religion" as a matrix, thus preserving the primacy of "religion" even when accepting its decomposition into heterogeneous functions. Such a way to put things gives "religion" the status of an original matrix, prior to what the researcher is trying to analyze: a fairly problematic hypothesis, both historically and heuristically, since "religion" could be an explanation only if we were able to define it precisely. However, all we can observe in the history of cultures are the implementations of various functions, that appear, grow and combine according to variable configurations, sometimes associated with a particular religion but sometimes also with non-religious entities; e.g., "belief in the supernatural" can be implemented in clairvoyance without being considered as properly religious.
In short: we should finish with "religion", and consider that religion is not an original matrix, but a contextual configuration. The Christian configuration is but the most familiar in Western societies, so that its specific functions spontaneously organize our conception of "religion" [13]. This configuration or, as Veyne says, this "form" ( [9], p. 486) is not universal but only, in some contexts, "majority" ( [9], p. 525). Moreover, in our context, it is the Catholic religion-a very particular configuration of features-which provides the paradigm of "religious" on which, wrongly, we tend to define and classify all kinds of heterogeneous phenomena. However, to relativize this paradigm, we need the "distant gaze" of the historian of antiquity or the anthropologist of exotic societies, able to splitting off "a Western, medieval and clerical thought" ( [35], pp. 243-44).
Going still a step further, we can argue that not only "religion" is not a substance but a mere contextual configuration: it is also an indigenous notion rather than a tool for scientific thought. So, comparing a phenomenon to religion is not achieving a heuristic, analytical, explicative operation, but a performative operation, that is, treating a problematic reality in an axiological way, be it to magnify the value "religion" or, at the opposite, to magnify the value "rationality". In short, it means standing on a moral or on a political level, but not on a scientific level.
Instead, heuristic comparison through functional decomposition allows a fine description of the observed phenomenon, without prejudice to its nature. It opens the possibility of an investigation that precisely departs what separates and what unites, for example, stars and saints: for, as stated by a French historian who has long explored this phenomenon, "the saints of Christianity were not the successors of the pagan gods. The stars of today do not follow the saints. They probably borrow some features from them, but they are part of a deeply different social system, or even 'religion'" ( [36], p. 17). Now, the relevance of analogy has found its limits, while opening the possibility of a rational comparison, that is to say, analytical and systematic, covering both differences and similarities. It thus becomes possible to speak not only of celebrity culture but of celebrity cult, without quotes, because we are no longer using a metaphor, but analyzing actual behaviors, whose specificities would definitely be missed if we reduced them to "religion".

Conflicts of Interest
The author declares no conflict of interest.