Is Atheism a Religion? On Socio-Anthropologic Cognitive Imperialism and Problems That Follow

The point of the joke lies in inserting non-religious people into a religiously contextualized frame. In this particular case, it revolves around Jehovah’s Witnesses’ evangelization strategies; also quite familiar in our Polish reality. The adherents’ door-todoor preaching serves to persuade others to their beliefs. The difference is that here we speak of atheists. As people who distance themselves from off types of religions, they should not be practicing any religious forms of evangelization. In the joke, the blank piece of paper symbolizes exactly the lack of a religious stand. It is symptomatic, since speaking of atheists as people who are religious in their own specific way and of atheism as something religion-like, and as paradoxical as it may sound at first, is more common than one might assume.


Atheism as a Religion-like Phenomenon?
There is a pictorial joke one can find on the Internet.
It is somewhat interesting in the context of my considerations. A man opens his front door to a couple of people distributing leaflets. They hand him one, but the paper is blank. The surprised man asks what this is all about, to which the two reply: "We are atheists." The point of the joke lies in inserting non-religious people into a religiously contextualized frame. In this particular case, it revolves around Jehovah's Witnesses' evangelization strategies; also quite familiar in our Polish reality. The adherents' door-todoor preaching serves to persuade others to their beliefs. The difference is that here we speak of atheists. As people who distance themselves from off types of religions, they should not be practicing any religious forms of evangelization. In the joke, the blank piece of paper symbolizes exactly the lack of a religious stand. It is symptomatic, since speaking of atheists as people who are religious in their own specific way and of atheism as something religion-like, and as paradoxical as it may sound at first, is more common than one might assume.
The idea of "religion-like" atheism occurs frequently in all sorts of discourse. In fact, it is so frequent that an attempt to catalogue its examples would yield a multivolume monograph. It will suffice that when Russell Blackford and Udo Schüklenk undertook the task of classifying the 50 most often recurring stereotypes (or as the authors describe them: "myths") on atheism, the idea of atheism as another type of religion took a primary position in their work (see Blackford and Schüklenk 2014: 18-24).
As far as the best known examples of such an approach are concerned, we must mention such important 20 th century Protestant theologists as Karl Barth, Paul Tillich or Harvey Cox, who broaden the scope of meaning of "what-is-religious" to the point of leaving almost no margin for a non-religious sphere to exist (see Demerath 1984: 363;Guja 2009: 145). It is here that one might seek sources of all subsequent examples of inventing religions or "churches without God," such as Sea of Faith launched by Anglican priest Don Cupitt (see 1998) or the French intellectualist (and a committed atheist) Alain de Botton's "Atheism 2.0" project. They both presume that a religion is a capacious enough form of social life that in the post-secular world it also incorporates atheists, who feel the need of ritual, connection and symbolism. We come across a similar approach in the last book by philosopher Ronald Dworkin entitled "Religion without God" (see 2014).
A slightly different reasoning to support the thesis on atheism's religion-like character abridges to pointing out that atheists too employ cultural codes and elements specific to religious organizations: they unite into communes (semi-churches); organize their own rituals (humanist weddings, dedication ceremonies or secular funerals); have their own symbolism (there is a number of such atheistic and freethought symbols -see Tyrała 2014b). There is even a demand for "atheist devotional objects" engaging that symbolism, especially in the form of T-shirts, mugs, bumper stickers et cetera (see Omyliński n.d.).
Indicating these types of formal similarities between atheism and religion often aims to criticize Is Atheism a Religion? On Socio-Anthropologic Cognitive Imperialism and Problems That Follow We have to resist pretending that actors have only a language while the analyst possesses the meta-language in which the first is 'embedded'. As I said earlier, analysts are allowed to possess only some infra-language whose role is simply to help them become more attentive to the actors' own fully developed metalanguage, a reflexive account of what they are saying. (B. Latour 2005: 49) T he purpose of this article is to present problems with defining non-religious people, as well as,the various dimensions of the consequences of the problems. Both theoretical and practical in nature, said problems stem from a long-lasting and still inconclusive dispute over how to define religion. I have come across these issues personally over the course of my research into non-religious people. I would thus not only like to sketch the dispute's framework, but also present how I have dealt with the problems. DOI: https://doi.org/10.18778/1733-8069.13.1.06 and devalue the former's claim to uniqueness and a dissimilarity to religion. It also causes atheists' resistance as they counter this attitude, regarding it as an indication of the incomprehension of the gist of atheism 1 . That in turn, breeds the aforementioned ethical problem condensed into the question of whether a (non-)religion researcher, or any other publically vocal member of society, has the right to label anyone against their will; even if such a researcher should be convinced he or she is sufficiently equipped with the theoretical tools to do so.
In this case, such a theoretical tool, considered by many religion sociologists as sufficient justification to treat a non-religion as a form of religion, is the functional strategy of defining religiousness.

The Debate over Defining Religion
Discussing the issues of defining religion, Günter Kehrer begins with the fairly pessimistic and discouraging statement that "none of the problems that religion sciences deal with has been as it is to resolve, the problem is worth devoting increased consideration, and it is worth doing so at the very start of any (non-)religion sociology deliberation. 2 I pre-1 In this article it is not my purpose to discuss a slightly different case of the so-called "atheistic religions" status. Michael Martin (see 2007), as much as he firmly self-declares as a supporter of treating atheism as something qualitatively completely remote from religion, he still recognizes the religious character of such religions acclaimed as atheistic as Jainism, Buddhism and Confucianism. He is conscious of various types of problems if not an inner contradiction of the oxymoronic statement that "atheistic religion" poses. This issue, however, goes beyond the premise of this article. 2 Not all agree with this approach to the matter. And so, according to Weber, "to define religion, to say what it is, is not possible at the start of a presentation such as this. Definition can be attempted, if at all, only at the conclusion of the study" (Weber 2002: 318). Eventually Weber never completed the presentation he mentioned with any definition of religion. Despite that, his profound analyses of world religions still inspire. As much as suppose that the definition of non-religion depends on the way religion is defined. Due to their dialectic relationship, the concepts of religion and non-religion are mutually conditional; and in the semantic sense as well.
The concept of religion resembles that of, for instance, culture. Each one of us uses them on a daily basis, in a way that is mostly intuitive and fairly adapted to a given context. But when it comes to providing their definitions it appears not so easy a task. It has caused, both religion sociologists and religious experts, headache; and still does. In the course of the evolution of those fields there have been many, often decisively divergent, definitions of religion. It would appear later that some of them have similarities, and so could be grouped into categories. We can then speak of the various strategies of defining religion. The most passionate and significant debate from the point of view of non-religion studies occurs between the supporters of two such strategies: the substantive and the functional defining of religion (Kehrer 1997: 19-32;Libiszowska-Żółtkowska 2004).
Those who support the substantive definitions of religion 3 focus on attempts to answer the question of what religion is. In this case the religion-defining criterion is its content. Typically, they try to isolate the factor that determines the essence of religion.
I seek those optimal definition strategies in this study, I do, however, realize that sociology of religion and non-religion may function despite a lack thereof. 3 Among them, to name a few, are Peter Berger (although his case is not indisputable; Irena Borowik brings attention to the sections of his idea where he happens to define religion also through the prism of its function -see Borowik 1997: 23); Steve Bruce; Roland Robertson; Karel Dobbelaere and Mircea Eliade.

Radosław Tyrała
Various kinds of entities, objects or types of reality: gods, deities, spiritual individualities, sacrum, sacred things, transcendence, meta-empirical reality, supernaturalism, final things, sacred cosmos, and meta-natural reality (Libiszowska-Żółtkowska 2004: 67). Another distinctive element of this religion defining strategy is a dichotomous view on reality. Usually, we come across a vision of the world split into sacrum and profanum; the sacred and the profane; the meta-natural and the natural; the meta-empirical and the empirical; the supernatural and the innate; where the former in each duo is always selfsame with a religious element. It should be noted that such an approach to defining religion is close to its common perception, at least in the Western culture, owing mostly to Christian influences.
Christianity fulfills all the requirements to be recognized as a religion in the substantive sense (due to the clear identifiability of such substantive elements as God, the sacred or the supernatural).
The substantive approach is not, however, devoid of weak points that have long exposed it to criticism. Formulating his own definition of religion, Émile Durkheim already criticized any attempts to define it through the prism of the concept of supernaturalism and the concept of deity. From today's perspective, we can acknowledge that these definitions appear to Durkheim as overly ethnocentric.
They do describe Christianity well, but can by no measure be applied to a number of non-European religions. And so, the concept of supernaturalism, as the French classic puts it, "it is certain that this idea does not appear until late in the history of religions; it is completely foreign, not only to those peoples who are called primitive but also to all oth-ers who have not attained a considerable degree of intellectual culture" (Durkheim 1990: 21). The scientific revolution conditioned dividing the world into the meta-natural and the "natural." With the revolution the category of "the natural order of things" was formed, which meant considering reality in the categories of laws and causality. As a result, the empirically experienceable and scientifically explainable "natural" realm was created.
Whatever did not fulfill the requirements and settled beyond this realm became "that-which-is-supernatural." Except, one must remember that the boundaries between the two have not been given once and for all. They shift, mostly due to advanc- ings (Luckmann 2006). From this perspective, pretty much any ideology could be regarded as a religion, including those that are "commonly" recognized as secular. This is where the fundamental influence of the strategy choice to define religion is brought to light. As Kehrer puts it: "choosing a definition is not so much a matter of personal taste. From the functionalist point of view Soviet Communism, for instance, appears to be a religion; from the substantive perspective, however, it does not reveal traits typical of a religion" (1997: 31). 5 Here we reach one of the most significant differences between the two defining strategies. Substantive definitions are generally exclusive; functional definitions tend to be inclusive. And on the one hand, that inclusivism determines the power of functional definitions, but on the other, it exposes its adherents to serious charges posed by "substantialists." Contrary to substantive definitions, functional definitions cope well with ethnocentric charges (Europocentric in particular) since they require 5 Poland has had book monographs analyzing Soviet communism in terms of a religious phenomenon published; "Para-religion Communism" by Marcin Kula (2003) and "Faith of The Soviet Man" by Rafał Imos (2007) are worth mentioning in particular. no supernatural entities (implicitly understood as analogous to those present in Christianity) to qualify a semantic system as a religion. This fact may attest to the superiority of functional definitions. The traditional approach to religion stems back from functionalism, distinguished religion and its functional alternatives (e.g., science or sport) that took over its functions, but were not religions themselves. Whereas Karel Dobbelaere (2003: 144) remarks that when we define religion in the functional manner, speaking of functional alternatives to religion proves impossible. For every such alternative will essentially prove to be some form of religion.

Toward an Optimal Definition
It is neither my task nor ambition to resolve which of the definitional strategies described above is better, assuming of course such resolution is at all possible. But, one should consider which of the strategies to define religion is more adequate in the case of research into non-religiousness and non-religious people. I tend to lean more toward the substantive definition for two reasons.
First reason: inability to use the functional definition. As mentioned before, according to the functional definition, any semantic system or ideology may be regarded as a religion, including those of a secular character. Following this train of thought, with no effort any ideology created to negate religion could be considered a religion as well. As Kehrer puts it: Would it then be impossible to adopt an understanding of religion broad enough to embrace any kinds of references to sacred things, or in other words, things of highest significance to the agent or agents. Through that we could reach a definition whose scope would make the mental construct of a society sans religion impossible (1997: 29).
In the sense of functional definitions then, non-religious people actually are religious, just in a different way. This mode of defining religion makes it downright impossible to isolate the non-religious as a separate social category, divergent in the quality sense from the religious. The "religious" and "sacred" is defined so broadly that any substantially non-religious viewpoints become pushed outside the socio-logical field of vision; and even if that should not happen, they must be re-conceptualized according to new, "pan-religious" rules. An extremely revealing example of such ideological conditioning within the sociology of (non-)religion is provided by Ryan Cragun and Joseph H. Hammer, who form a thesis on the "Pro-Religious Hegemony in the Sociology of Religion" (see 2011). According to them, religion and its manifestations are the main 6 I have not conducted systematic analyses of the religious standpoint or denomination adherence of sociologists declaring for various types of religious definitions, but I suppose that the functional and inclusive definitions supporters will more frequently be religious people (one can treat such statement as a hypothesis introduced for the sake of possible future research into the issue). The fact brought up at the beginning of this article attests to that; many of those mentioned who support defining atheism as a "religion-like" phenomenon are clerics or theologists. On the other hand, substantive definitions are generally used by non-believers, including publications by social movement activists. frame of reference (axiological as well) for sociologists of religion. For the most part, sociologists of religion, more or less consciously, represent a church denomination, and so,an ideologically entangled point of view. Thus religion and the state of "being religious" seems to them as some (implicitly desirable) norm, and anything beyond that norm turns out to be a deviation. According to the authors, it shows up primarily in the terminology used by sociologists of religion. As much as the terms pertaining to religiousness or substituting one religion for Because a sociologist thrives on seeking and explaining divergences present among people, from the cause-effect point of view, he or she does not lean toward defining seeming differences as philosophical similarities nor toward rejecting a common understanding by turning to sophisticated erudition. To the extent in which a believer and a non-believer think they differ, that difference is sociologically real, no matter what theology has to say (Demerath 1984: 363).

Non
The issue of differences between sociological and philosophic-theological strategies of defining an atheist or non-believer comes forth.
Also according to James Beckford (2006: 35-57), a supporter of a constructionist approach to religion, researchers should not attach themselves to a specific definition of religion. He begins with an observation that in the course of history religion has taken on a countless variability of forms. He detects a lack of consensus on the definition of religion in contemporary societies in particular. Every definition has a normative character (and thus will cause somebody harm). Beckford notices that even within one society, collective subjects such as legal institutions, the state, mass media, schools or healthcare facilities create their own concepts of religion and use them for their own purposes. This leads the author to the following conclusion: "It is better to recognize that a universal definition of religion is unattainable" (Beckford 2006: 46). Instead of continuing the aimless search, he proposed his own solution: 159-178). 8 In their eyes, the very rejection of religion perceived in substantive terms makes them, non-believers. That is who they believe they are and one can presume that they live according to such a self-definition. I then decided that adopting such a presumption will optimize my own research into non-believers functioning in the context of Christian religious culture. I claim that the culture that representatives of the communi-8 One might find an additional confirmation of the fact that non-believers perceive religion exactly in substantive categories in, say, the "Religion" entry (Stopes-Roe 2007) of "The Encyclopedia of Unbelief," a monumental work edited by Tom Flynn (2007). It aspires to be a kind of main reference work, written on non-belief from the point of view of non-belief. In the entry, religion is defined solely via a reference to supernatural entities, especially a God in a personified form. It is then a definition of remarkably substantive character, served in the form that nearly 100 years ago Durkheim submitted for criticism. ty under my scrutiny come across determines the choice of this particular definition instead of any a priori of terminological arrangements.

Research clues
In the text, I refer to my research entitled Conducting research into non-religious people in Poland, I used the terms "non-belief" and "non-believer." 10 I regarded every person that declared themselves a non-believer as such. A mere self-declaration was the necessary and sufficient condition for non-belief. 11 I did not create any a priori lists of requirements a person should meet. I did not ask any filter questions, for instance, if the respondent actually believed in God if they joined any religious rituals; if they followed a religiously grounded morality or rather some sort of autonomous ethics. The respondents would individually add meanings to the "non-believer" label. In my opinion, this self-declaration is the way to define non-religiousness in a form that is non-invasive toward self-definitions of the researched non-religious people. Thanks to such a definition strategy, we do not impose our own preferred conceptualization onto anyone and can follow the respondents' reflection. We also avoid potential conflicts that may stem from labeling non-religious people religious. It often raises their negative emotions that may result in 10 A detailed justification of the reasons that caused me to choose this particular term (and not one of many already existing ones such as: atheist, agnostic, rationalist, humanist, irreligious person, etc.) can be found in my studies (see Tyrała 2015). 11 I am not a pioneer on the Polish ground. The strategy of defining non-believers via their self-declarations was already in use in the Polish sociology in the 1980s by Mirosława Grabowska and her team (Grabowska 1990: 57). a number of inconveniences in the research process (reduced inclination or refusal to participate, or even various forms of aggression addressed against the researcher). The reason that I myself have not come across such violent reactions is likely partially due to the fact that I did not enforce any narrow definition framework, but instead allowed the respondents a vast freedom in describing and interpreting their own non-religious sub-worlds. Had I preliminarily assumed that the non-believer should be free of such beliefs, then those types of respondents would never be included into my sampling. It turned out, however, that not only did they make it in but they also comprise nearly one fifth to that 3.6% of survey respondents, declared non-believers -stated that they believe in a personified God and another 5.6% stated they had difficulties in determining who they believe in. In spite of these declarations seeming contradictory to any known to me, common or "ex- under the same roof. But with progressive changes in 12 Emil Durkheim can be thought the "father" of such approach in sociology. In "The Rules of Sociological Method" he exhorted to studying sociological facts "from an aspect where they appear separate from their individual manifestations," which one could understand as the deed of foundation for treating scientific definitions within sociology as superior against the common one; the supremacy of etic over emic (see Durkheim 2000: 74-77). the family shape and general approach to family life, the model turned out ever less adequate. For instance, according to that approach, homosexual couples raising children were not considered families. According to that approach your family dog or cat could not be considered a family member in spite of being frequently treated as such by their human cohabitants, as research shows (see : Konecki 2005). Due to all those dilemmas, in the 1990s Jon Bernardes adopted the Thomas theorem for family research needs (later even described as the "Bernardes theorem"). The new version states: "we must entirely reject the concept of 'the family' as created in our social sciences, as theoretically inadequate and ideologically engaged, and as a consequence of the rejection research only into what the actors of social life will themselves consider a family and family life" (Bernardes 1993: 40;see Szlendak 2011: 105). Despite the facts that this approach has not been employed by all sociologists of family, it has not been free from criticism and it is not an ideologically neutral approach, one must admit that it does solve certain significant aporias present in the field. It has encountered resistance mainly among sociologists of a more conservative attitude, professing a more traditional vision of the family, prophesying its crisis, uninclined to expand its definition enough to include homosexual families or house pets. Just as it does on the religious field, it distinctly shows that definition debates rarely have an ideologically neutral character and can be interpreted as a type of game played by representatives of the various interest groups, often conflicted as far as a general outlook, politics or religion are concerned.
In both cases, research into the family and the example discussed in this article, following the meanings ascribed by studied actors (in both cases amounting to using self-declarative strategy) solves the problems that alternative solutions do not handle (although, as I have shown, they incidentally generate new ones). In my opinion, in the case of research into non-believers this approach turned out most effective because by using it I discarded the most perturbing research dilemmas in the three significant dimensions of my analyses: epistemological, conceptualizing and ethical.
In the epistemological sense following the actors, that is, non-believers applying a substantive understanding of religion, means that irreligion becomes "recovered" and it even becomes possible to consider and scientifically articulate it. As I have mentioned before, using functional definitions of religion trims irreligion at its epistemological roots since they are definitions inclusive enough to automatically throw it in the "thatwhich-is-religious" bag. And that takes away its autonomy. Such mode of defining religion prevents isolating non-believers as a separate social category qualitatively divergent from religious people. In the conceptualizing sense, following the actors solves the dilemma of defining non-religious people for research purposes.
It, so to say, inflicts a self-declarative strategy as the optimal one instead of the strategy that predefines non-religious people by using a list of markers that the researcher determines in advance. Despite, as I have mentioned, not being problem-free, this approach still yields exceptionally interesting research results. Finally, following the actors turns out to be the optimal approach in the ethical sense. By using it we protect the respondents against a situation in which they are restrained by the power of scientific authority to a definition with which they do not identify and which often serves as a negative framework. By using it we make the researcher-respondent interaction more symmet-rical. It is for those three reasons I have made every effort to avoid treating non-believing respondents as religious people. With passing time, I still view it as the optimal solution.