So far, all of the philosophy of religion I have read focuses on the questions of God's existence and on the problem of evil. It seems to me that just as important would be the question of what a rigorous definition of religion is. It seems to me that there should be a demarcation problem for religions similar to the one Popper, Feyerabend and others worked on for science.

  1. Have philosophers addressed the questions of what constitutes a religion and what doesn't?
  2. How to differentiate between a secular ideology or philosophy and a religion (for example the status of Confucianism or some Schools of Buddhism)?
  3. How to differentiate between a cult (in the heaven's gate sense of the word) and an established religion? What exactly gives a religion or denomination accepted status?
  • 4
    While I see the thread connecting all of these questions, as a whole, I think they are too large to answer in a single SE question and that some of them are off-topic (i.e., #5 (largely opinion-based for philosophy -- better for sociology), #4 (more a question of political systems than philosophy proper unless you want to ask whether this is ethical which is quite different than the conditions where it happens), and part of #3 (cult has a sociological definition and a common parlance one)).
    – virmaior
    Jun 17, 2015 at 0:17
  • If I am to ask them separately, I'm afraid some people will vote them as duplicates. Jun 17, 2015 at 0:52
  • As for #4, I see it as an ethics question, not a political one - I am assuming a rational, secular and liberal society (as opposed for example to a theocracy which defines religions based on the scripture of the state religion). Jun 17, 2015 at 0:54
  • 3
    I agree with virmaior that this is overstuffed, 3),4) sound like questions a lawyer would ask when applying for tax exempt status, even if the intent is different, and 5) is dictionary definitional en.wikipedia.org/wiki/Organized_religion The philosophical part of the question is whether there is a philosophical test for religion like (there seemed to be) one for science a la Popper. But it seems to me that religions, even more so than languages, illustrate Wittgenstein's family resemblance: plenty of overlapping similarities, no universal features.
    – Conifold
    Jun 17, 2015 at 3:20
  • 1
    I found Idinopulos's essay illuminating crosscurrents.org/whatisreligion.htm "We don't exactly know what we mean by the word, religion. We don't know how to use the word or what constitutes a misuse of the word. It would be convenient to assume that by "religion," we mean the fetishisms, animisms, polytheisms, and monotheisms of the known historic religions... that all the religions were like branches of a large tree, with a visible trunk... If... the diversity of religions is taken seriously, we would not think that religions are branches of a single tree".
    – Conifold
    Jun 18, 2015 at 1:20

6 Answers 6


Humans love to categorise things, and we may think of categories as dividing things up with borders between the categories. But that isn't actually how we normally do conceptualise our categories - instead we categorise things according to their likeness to archetypes or prototypes, the central most typical examples of a category. This is called Prototype Theory. The borders between categories are usually extremely hard to determine, but the centres are easy to recognise.

We know what the prototypical religions are: Christianity, Islam, Hinduism, the ancient Greek pantheon, etc. We know what some things that are most definitely not religions: environmentalism, egalitarianism, communism, anarchism, etc.

This leaves some things that are harder to classify, but that always happens when you classify according to prototypes rather than borders. There's a trade off: prototypical classifications are easy to define but it's harder to classify the cases which fall between prototypes, whereas border classification systems are hard to define, but it's easy to classify things once you have come up with a definition. I think that prototypical systems should be preferred though as they are closer to our natural conceptualisations, and border classification systems inevitably come to categorise something on one side of the border which intuitively feels like it belongs on the other, so you either have to change your border definition, or you deal with a system that's partially broken.

  • 1
    I'm not so convinced "we know some things that are most definitely not religions". For communism, see Hitchens youtube.com/watch?v=J7Xpa2RgBjo . For environmentalism, there's some Gaia-religion versions and the basic one has a concept of sin and redemption. Early research in religion (the academic discipline) was templated on Christianity which extended pretty easily to Islam and Judaism, but I think the current position is that it's really hard to clearly state something is not a religion.
    – virmaior
    Jun 19, 2015 at 1:59
  • 2
    @virmaior Just because people have taken the ideas of shared ownership of production (etc) and care for the environment in religious directions does not mean those core ideas are religious. The very fact that they have different names (like the Gaia movements) means they are referring to different movements. There are some things which are hard to classify, but there are still many things which are easy to classify. Democracy is not a religion. Mathematics is not a religion. Fried potato is not a religion. Jun 19, 2015 at 2:05
  • 2
    I applaud the optimism, but "Theravada Buddhism is usually included in any book on the world's religions, it is not theistic, recognizes no sacred being or beings, and does not officially encourage worship of Buddha or any "higher being" (despite popular veneration of the Buddha-ideal). Theravada Buddhism appears to be a technique or program for human self-purification or self-fulfillment or self-negation". crosscurrents.org/whatisreligion.htm Scholars question not just the possibility of any classification of religions, but even its usefulness.
    – Conifold
    Jun 19, 2015 at 2:30
  • 1
    @Conifold I don't see how that is any evidence against my answer - Buddhism and Confucianism are the hard to classify cases that are not close to the prototypes. Please look up Prototype Theory if you are not very familiar with it. Usefulness is another matter, which the question does not address. Jun 19, 2015 at 2:42
  • 1
    Prototype Theory means that just because some things are hard to categorise, it doesn't mean that the categorisation system is useless or impossible. We should instead expect that all categorisation systems will have edge cases that are hard to categorise, because the basis of the categorises is their centres, not their borders. Jun 19, 2015 at 3:18

It's worth noting the etymology of the term you are looking to demarcate. Per "The Etymology of Religion" by Sarah F. Hoyt (1912) a commonly understood root of religion translates from the Latin religare which means "to bind". Also of note, relegere from "re-" (again) and "legere" (to read). The word also comes to us from religio (in the sense of obligatio) meaning "obligation, bond, reverence" where "re-" is intensive.

I think at best you can assert a general notion, but the edge and corner cases may not be adequately addressed. For such, you might want to look at legal understandings of cults (which kinda boil down to religions that engage in illegal activities) and the registering churches or establishing of religions. In the general sense, I think community, ritual and reverence cover religion. The legal definition of what is and is not a religion is ongoing in the U.S.

For some, this ambiguity might be problematic, for example:

But the word, without referring to its etymology, has, in the manner it is used, no definite meaning, because it does not designate what religion a man is of.

Thomas Paine, "On The Meaning Of The Word Religion, And Other Words Of Uncertain Signification"

...however, I don't think it is a problem for philosophy. Considering that what you are trying to demarcate is a human activity, I don't think you can advance by the same means that Popper distinguished the scientific and the pseudo-scientific. The institutional fact (per Searle's "Construction of Social Reality") of an established religion is, after all, a matter of agreement and, appropriately enough, deontic powers.

As for distinguishing a religion from a secular ideology, well, a secular ideology is just non-religious belief, however, there is nothing preventing the establishment of a religion upon a secular ideology.


Towards question one, In 'The Essence and Effect of Religion' Bertrand Russell defined religion as

"...beliefs with many dogmas which direct human behaviour and are neither based on—nor contradict—real evidence; and that the method employed by religions to direct people’s minds are based on sentiment or power rather than reason."

Most or Russell's work on religion is regarding Christianity, but I think the definition above work even for Confucianism and Buddhism.

The interesting thing about this definition though is that it could encompass almost all beliefs held by the majority of people (who, according to considerable weight of sociological study, derive their beliefs from others rather than rational analysis)

To be religion then I think a further caveat needs to be added that the members of that sect openly admit that their beliefs are thus derived rather than claim them to be the result of rational analysis. This would only apply in the modern era as it would rule out a considerable amount of Medieval religious philosophy aimed at proving the existence of God, but that the definition of religion has changed seems to be entirely consistent with the definition of many other words.

The second question I think has been perfectly adequately answered by @curiousdannii above.

Question 3 is a sociological one. At various times in history, religious sects which are today mainstream have been considered cults so the term is a derogatory one used to denigrate a particular group's beliefs. It's application is therefore not to do with ontology (in the information sense of the word), but social motives.

  • I agree, but the modern case is not quite so simple. Many religious people say, read their book & you will be convinced too, not 'we accept this stuff was just made up'. Scientists don't have capacity to check each other's results all the time, they rely on a network of trust and authority. As do we all non-scientists, who rely on their results.
    – CriglCragl
    Mar 2, 2018 at 10:54
  • Russell's definition would have to include any system of morality, or culture generally (eg table manners) as religious. It is too broad.
    – CriglCragl
    Aug 6, 2021 at 13:35

Definition's like Russell's are religion by assumption, rather than observation. Religions have no requirement to be developed from axioms, or core tenets. Perhaps most crucially, there aren't means to reform and update them universally accepted by their adherents. Comparison of Islam's schools of jurisprudence, rabbinic scholarship, and Christian councils and sect governance, is crucial to understanding how these religions got where they are. Especially interesting, as they are all part of one religious culture.

I go with the argument presented here https://aeon.co/ideas/whence-comes-nihilism-the-uncanniest-of-all-guests that religion performs primarily a different function than presented in these definitions by assumption. That is, a kind of metis, a crafty way, of trying to live meaningful lives together. But, a previous eras metis becomes 'illegible' as modes of life and challenges change. Which is where reform and development mechanisms come in.

I am a flag waving agnostic. That is, not a Russellian agnostic-atheist, a stance of being scientifically agnostic but culturally anti-religion. I see the value of agnosticism as around, not dismissing out of hand all of the 'moth eaten musical brocade' of religion. We have come culturally to see ourselves as purely individuals, with no inheritances, and no real stake in the future beyond our own desires and needs. This is unravelling things like secularism as the cure bitter identity politics, and the ecology and biodiversity of the world that will need sacrifices on our part to save.

Religions have helped us live together, and assert lived values which have benefited us all. And then, helped transmit those values, between generations, eras, even epochs. Philosophy builds a corpus of texts which are there for those interested. It doesn't demand transmission as duty, or even any more to be culturally refined or a good citizen. As postmodern antifoundationalism leaves us fragmenting into ever smaller and more selfish cultural units, we are losing the skills to assert and transmit piditive values and ways to be.

So I am saying there is a religion-shaped hole in our society. There are passionate advocates of filling that gap with science, but they actually get very polarised and bitter even with each other, as they try to take science beyond it's domain. I think we need something more like the aesthetics-cultural embodiments found in Buddhist-infused Japanese tea ceremonies, where architecture, poetry, pottery, socialising, and mode of living, are brought together. We need to advocate not just a body of ideas, but of literature and poetry and architecture, capable of reform and transmission, so that our community values can develop rather than slide into illegibility and fragmentation.


Getting in touch with a higher power is religion. So, religion seems to me an act more than a stance whereas an ideology could be characterized as a position, or stance. One can't be religious and not practice their religion.

The actions of a cult lead a discord between the end of their means and reality, typically manifested as a rejection by society.

  • 1
    AS mentioned in the comments above, neither Theravada Buddhism, nor Confucianism have higher powers. Yet both are considered religions. Jun 19, 2015 at 4:39

In Russia there is a term "destructive sect" or "destructive cult". While many new religions often get labeled as such, there are people who are trying to give a strict definition of a "destructive sect".

Particularly, some of the following criteria were suggested:

  • Anti-family attitude. The followers are getting indoctrinated against their relatives and parents.

  • Strict demands for donating big sums of money.

  • Propaganda of suicide and/or predictions for soon end of the world.

  • Use of drugs in the indoctrination techniques.


  • Modern China has a big problem with radical religious movements, like Falun Gong. Chinese history has been shaped by such movements, like the Boxer Rebellion, and most of the unrest that led to civil wars and change in dynasties, which mostly originated with religious-cultural movements. Demarcation is clearly going to be different for different cultures.
    – CriglCragl
    Mar 2, 2018 at 10:50
  • This post does not answer how science or politics, for example, is different from religion. It might answer p.3, but there is a huge difference between destructive cult and cult in general which are not all called religions.
    – rus9384
    Nov 7, 2018 at 8:42

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .