Anthony Campbell

Religion and Language

Last revised 21-09-2020


Religion isn't so much what you believe, it's the stories you tell. But stories need words (well, they can use pictures too, but words are needed to explain the pictures properly). So there's a trivial sense in which religion and language are related. You couldn't acquire a religion without using language. But the connection goes deeper than this. Both religion and language are closely connected at another level and are acquired in quite similar ways.


Something we would recognise as a religion exists in practically all the societies studied by anthropologists. Some people have interpreted this universality as indicating the presence of a "religion instinct", an inbuilt tendency to religious belief and practice in all human beings. Tbjs in turn implies the existence of brain structures that give rise to religious beliefs.

Very similar arguments have been applied to language. Every human society we have encountered has possessed language, and Noam Chomsky has famously claimed that there are similarities in the structure of all languages that point to the existence of a "universal grammar" (Chomsky 1972). The grammar or "deep structure" of human languages is complex, yet young children seem to have an innate ability to master this complexity within a short time, as if by instinct. This has suggested to many people that the rules of grammar are in some sense built into the human brain during evolution.

If this idea is correct, the same may be true of religions. Perhaps there is a "deep structure" for religion just as there seems to be for language. But is it correct?

Is Chomksy right?_

In his book The Symbolic Species Terrence Deacon rejects Chomsky's view and proposes instead the hypothesis that languages evolve in a kind of symbiotic relation with the human mind (Deacon, 1997). The fact that young children are able to learn languages with apparent ease, he suggests, doesn't mean that they have some extraordinary innate linguistic ability but rather that human languages have evolved to be learned easily by immature minds.

There is a two-fold evolution going on here: certainly the human brain has evolved linguistic capabilities that are absent in the brains of other primates, but at the same time languages have adapted themselves to be readily learnable by children. This recalls Richard Dawkins's meme idea, which Deacon does mention in passing; Dawkins relates this to the spread of religion in human societies.

Resemblances between language and religion

If we now look at religion we find that Deacon's view of language applies quite well to that. Like language, religion has evolved to be easily acquired by children. Consider the following.
  1. Religion and childishness

    Religious people are often reproved by the non-religious, and even by some co-religionists, for having a childish view of God; and this is in a sense reflected in references to God the Father (or sometimes nowadays God the Mother). If religion has evolved to be easily learned by children this makes good sense. "Except you be converted, and become as little children, you shall not enter into the kingdom of heaven." (Matthew 18, 3).

  2. Conversion vs early-acquired religion

    The language-learning ability of children is different from that of adults. There is a long-held view that this indicates a critical period for language learning, similar to the imprinting phenomenon in birds (newly hatched chicks follow the first moving object they see, even if it's a human being). Deacon's view is rather that a degree of immaturity may be actually necessary for language acquisition in this way.

    Whatever the explanation, the phenomenon certainly exists, as anyone who has tried to learn a new language in later life can testify. But religion is acquired by children in a very similar way to language. Many people are taught religion literally at their mothers' knees, and religions infused early in life in this way have a different quality from those that may be adopted later as the result of conversion.

    Religious beliefs inculcated in childhood are also difficult to shake off, just as one's mother tongue is more persistent in the face of disuse than languages learned in later life. Seen in this way, the well-known if apocryphal Jesuit saying, "Give me a boy until he's seven and he's mine for life", takes on a new significance.

  3. The language of religion

    Acquiring a religion may involve learning a new vocabulary and syntax: for example, the old Quaker use of "thee" and, in some Christian circles, phraseology such as "believing on Jesus" instead of the vernacular "believing in". Those who adopt these ways of speaking subtly distinguish themselves from non-believers, rather as having a regional accent acts as a linguistic badge of origin.

  4. Sacred languages

    Many religions have a sacred language (Hebrew for Judaism, classical Arabic for Islam, Sanskrit for Hinduism, Pali for Theravada Buddhism). Because religions are generally ancient the languages they use are often partially or wholly unintelligible to the laity and sometimes even the clergy, but contrary to what religious modernisers suppose, this linguistic remoteness is a strength not a weakness. Misguided attempts to bring the language up to date often coincide with a loss of religious confidence, and it's diffcult to separate cause and effect. Some Roman Catholics still lament the abandonment of the Latin Mass in favour of the vernacular; disuse of the Book of Common Prayer by the Church of England has not prompted an influx of young worshippers to the pews (Freeman 2001).

  5. Dialects in language and religion

    Over time, languages acquire regional dialects, and the same is true of religion. There is a tendency for two separate trends to form within mature religions, one plain, the other ornate.

    (i) In Christianity we have Catholicism and Protestantism: Catholicism goes in for devotion to the Virgin Mary and the saints and produces complex vestments and rituals, all of which are frowned on to a greater or lesser extent by Protestants.

    (ii) In Buddhism there is the distinction between Theravada and Mahayana: Theravada is relatively austere and unemotional, whereas Mahayana has the Bodhisattvas (Buddhas-to be, who compare in some ways with the saints in Catholicism) and elaborate ceremonies.

    (iii) Within Islam there are differences in tone between Sunni and Shia: in a Shia country such as Iran you frequently see pictures of Ali, Husayn and other "saints" in taxis and elsewhere which are reminiscent of Greek icons and Catholic holy pictures.

    Catholicism, Mahayana, and Shiite Islam have something in common, and so do Protestantism, Theravada Buddhism, and Sunni Islam. We can think of these as religious dialects. They mostly remain mutually intelligible, at least to start with. But over time the process may continue to a point where that is no longer true, and then we can speak of the evolution of a new religious species.

  6. Speciation in language and religion

    Deacon describes linguistic evolution as follows.
    As a language passes from generation to generation, the vocabulary and syntactical rules tend to get modified by transmission errors, by the active creativity of its users, and by influences from other languages... Eventually words, phraseology and syntax will diverge so radically that people will find it impossible to mix elements of both without confusion. By analogy to biological evolution, different lineages of a common ancestral language will diverge so far from each other as to become reproductively incompatible.
    Substitute "religion" for "language" and "ritual" for "syntax" in this passage for a pretty exact description of how Christianity evolved from Judaism. They have become different species, which can no longer interbreed. But this is a matter of degree. In biology there may be subspecies which are capable of interbreeding although they seldom do so. The different Christian denominations can be thought of as subspecies of the parent religion.

  7. Did language and religion originate together?

    Finally, and very speculatively, the origins of both language and religion may go back to the very beginnings of modern human consciousness. Many people believe that there was a qualitative shift in consciousness about 50,000 years ago—the Great Leap Forward, when tool-making became more complex and the cave paintings in France and Spain first appeared. We don't know why these paintings were made but a prevalent idea is that they had some sort of religious or magical significance. We also don't know when language first developed, but some researchers think that an elaborate form of speech first became possible to humans at about the same time as the paintings were made. If these ideas are at all correct, it would follow that language and religion were closely connected at their very inception. (Recent archaeological discoveries suggest much older starting date for art and language but the point still stands.)

Religion: parasite or symbiont?

According to Deacon we can think of languages as parasites or viruses. But that is probably too severe, as he concedes, since languages are after all beneficial to their hosts and should therefore better be regarded as symbionts—organisms that live together for mutual benefits. Is that the right way to think about religion?

We couldn't do without language, but could we do without religion? Has it become so deeply infused into our minds and our culture that we cannot rid ourselves of it? It may be like the mitochondria in our cells. These were originally free-living organisms, but at some stage in the distant past they became permanent denizens of all advanced cells, which depend on them for their ability to use oxygen for energy. Have religions become our psychological mitochondria?

Postscript added 9 February 2006

The late Ben Cullen of the Department of Archaeology and Palaeoecology, Queen's University, Belfast, wrote a paper shortly before he died called Parasite ecology and the evolution of religion. In this he criticised Richard Dawkins's view of religion as a parasite. Here is an abstract of the paper.
It is argued that the blanket view of religion as a disease, advocated by Dawkins, is inconsistent with the principles of parasite ecology. These principles state that vertically transmitted parasites evolve towards benign, symbiotic states, while horizontally transmitted parasites increase their virulence. Most of the world's established religions are transmitted vertically, from parents to children, and are therefore expected to be benign towards their hosts. Yet, certain horizontally transmitted cults, such as the Aum Shinrikyo, seem to effectively exploit their hosts in a way similar to an infectious disease.
To which I would add that many of the recent Islamic terrorist attacks have been perpetrated either by converts to Islam or by people who are described as having been lax in their religious observance before becoming radicalised. In both cases their recent religious views were acquired mainly or entirely by horizontal transmission.

Cullen's idea fits well with the view of religion which I propose in this article: namely, that it can be either beneficial or harmful to its host (or possibly neutral). Most of my discussion concerns vertical (parents to children) transmission, which would generally be beneficial or neutral.

See also Religion as Parasite, Parasite as Religion.