Like it or not, we are surrounded by religions. The 'Great Faiths' of the world have lasted thousands of years and affect our calendars and holidays, our education and upbringing, our beliefs and our morality. All over the world people spend vast amounts of time and money worshipping their gods and building glorious monuments in which to do it. We cannot get away from religions, but using memetics we can understand how and why they have such power.

All the great religions of the world began as small-scale cults, usually with a charismatic leader, and over the years a few of them spread to take in billions of people all across the planet. Imagine just how many small cults there must have been in the history of the world. The question is why did these few survive to become the great faiths, while the vast majority simply died out with the death of their leader or the dispersal of their few adherents?

Dawkins was the first to give memetic answers (Dawkins 1986, 1993, 1996b), although his ideas on religion have frequently been criticised (Bowker 1995; Gatherer 1998). He took Roman Catholicism as an example. The memes of Catholicism include the idea of an omnipotent and omniscient God, the belief that Jesus Christ was the son of God, born of the virgin Mary, risen from the dead after his crucifixion and now (and for ever) able to hear our prayers. In addition, Catholics believe that their priests can absolve them from sins after confession, the Pope literally speaks the word of God, and when priests administer the mass, the bread and wine literally change into the flesh and blood of Christ.

To anyone uninfected with any Christian memes these ideas must seem bizarre in the extreme. How can an invisible God be both omnipotent and omniscient? Why should we believe a two-thousand-year-old story that a virgin gave birth? What could it possibly mean to say that the wine 'literally' becomes the blood of Christ? How could someone have died for our sins when we were not even born? How could he rise from the dead, and where is he now? How could a prayer, said silently to yourself, really work?

There are many claims for the efficacy of prayer in healing the sick, and even a little experimental evidence (Benor 1994; Dossey 1993), but few of the experiments have controlled adequately for placebo effects, expectation, and spontaneous recovery, and some have shown that people with the strongest religious faith were less likely to recover from acute illness (King et al. 1994). Against the claims are hundreds of years of people praying for the health of their royal families or heads of state with no apparent effect, and the inability of modern-day religious healers to make any obvious difference in hospitals. Then there are all those countless wars in which both sides routinely pray for God to help their side and kin the enemy. Yet millions of people all over the world profess themselves Catholics and pray to Jesus, his mother Mary, and God the Father. They spend vast amounts of their valuable time and money supporting and spreading the faith to others, and the Catholic Church is among the richest institutions in the world. Dawkins (1993) explains how religious memes, even if they are not true, can be successful.

The Catholic God is watching at all times and will punish people who disobey His commandments with most terrible punishments - burning forever in hell, for example. These threats cannot easily be tested because God and hell are invisible, and the fear is inculcated from early childhood. A friend of mine showed me a book he once treasured as a child. It had pictures of a little good boy and a little bad boy. You could open up the flaps of their blazers and inside the good boy find a white and shining heart, while the bad boy had a black spot for every sin he had committed. Imagine the power of that image when you cannot see inside your own body and must only imagine the little black spots piling up and piling up - when you talk in class or cheat in a test, when you take your sister's toy or steal a chocolate biscuit, when you think a bad thought, or doubt God's truth and goodness ... every one a black spot.

Having raised the fear, Catholicism reduces it again. If you turn to Christ you will be forgiven. If you honestly repent of your sins, bring up your children as Catholics, and go regularly to mass, then, even though you are unworthy and sinful, God will forgive you. God's love is always available but at a price, and that price is often overlooked completely because it is paid so willingly. It is the price of investing massive amounts of time, energy and money in your religion - in other words, working for the memes. As Dawkins pointed out, Catholics work hard to spread their Catholicism.

I previously described several meme tricks that New Age memeplexes use. All these can be found in religions too. First, like alien abduction and near-death experience memes, religions serve a real function. They supply answers to all sorts of age-old human questions such as: Where do we come from? Why are we here? Where do we go when we die? Why is the world full of suffering? The religious answers may be false but at least they are answers. Religious commitment may give people a sense of belonging, and has been shown to improve social integration in the elderly (Johnson 1995). Religions may also incorporate useful rules for living, such as the dietary laws of Judaism or rules about cleanliness and hygiene which may once have protected people from disease. These useful functions help carry other memes along.

The truth trick is liberally used. In many religions, God and Truth are virtually synonymous. Rejecting the faith means turning away from Truth; converting others means giving them the gift of the true faith. This may seem odd when so many religious claims are clearly false, but there are many reasons why it works. For example, people who have a profound experience in a religious context are inclined to take on the memes of that religion; people who like or admire someone may believe their truth claims without question. At the extreme, people will even tell lies for God and manage to convince themselves and others that they do so in the name of truth - as when 'Creation Scientists' proclaim 'The Truth' that the earth is only six thousand years old, and back it up with denials of the fossil record, or claims that the speed of light has slowed since the creation so as to give the illusion of a vast universe and an ancient planet (Plimer 1994).

Beauty inspires the faithful and brings them closer to God. Some of the most beautiful buildings in the world have been constructed in the name of Buddha, Jesus Christ, or Mohammed. Then there are the beautiful statues and alluring stories in Hinduism; stained glass, inspiring paintings, and illustrated manuscripts; uplifting music sung by tremulous choir boys and vast choirs, or played on great organs. Deep emotions are inspired to the point of religious ecstasy or rapture which then cries out for - and receives - an explanation. The ecstasy is real enough, but from the memes' point of view, beauty is another trick to help them reproduce.

The altruism trick permeates religious teachings. Many believers are truly good people. In the name of their faith they help their neighbours, give money to the poor, and try to live honest and moral lives. If they are successful then generally people come to like and admire them and so are more inclined to imitate them. In this way not only does good and honest behaviour spread, but the religious memes that were linked to that behaviour spread too. Alongside this comes merely the semblance of good behaviour. Hypocrisy can flourish when goodness is defined not only as kind and altruistic behaviour, but as sticking to the rules and obligations of the faith. Much of the money donated to churches, temples, or synagogues is not used for the poor or needy, but to perpetuate the religion's memes by erecting beautiful buildings or paying for clergy. Activities that spread memes are also defined as 'good' even though their benefit is questionable, such as saying prayers at specified times, saying grace at every meal, and keeping one day of the week as a day of worship. In this way huge chunks of every believer's time are willingly devoted to maintaining and spreading the faith.

Many people think of Mother Teresa as a saint. Indeed, she may soon be officially sanctified by the Catholic Church. She is many people's idea of the truly selfless and altruistic heroine. But what did she actually do? Some of the inhabitants of Calcutta accuse her of diverting attention from the real needs of the city's poor, of giving Calcutta a bad name and of helping only those who were prepared to take on Catholic teachings. Certainly, she was fiercely anti-abortion and anti-birth-control. Many of the people she helped were young women with no access to contraceptives, little ability to avoid being raped, and almost no access to health care if they became pregnant. Yet she steadfastly maintained her Catholic opposition to the one thing that would have helped them most of all control over their own reproductive lives. Whatever we may think about how much she really helped the starving people of Calcutta there is no doubt that her behaviour effectively spread Catholic memes by using the altruism trick.

Even evil and cruelty can be redefined is good. The Koran states that it is good to give a hundred lashes to an adulteress and to have no pity on her. You might think that Muslim women can avoid this by not committing adultery, but Warraq (1995) explains in unpleasant detail what life can be like in countries that adhere strictly to Islamic law. Women may be powerless to resist sexual abuse, and afterwards must take the punishment while the men who abused them get off free. Since women are objects of disgust, a man is supposed not to touch a woman he does not have rights over. Women are routinely locked away and, if they are allowed out, must walk behind the man and be suitably covered - which in many countries means being covered head to toe in a smothering garment with just a tiny grille to look out of. Obeying such rules to the letter makes a Muslim 'good', regardless of the misery it creates.

Returning to more honest uses of goodness and altruism, Allison's (1992) theory of 'beneficent norms' applies especially well to religions. One of his general rules is 'Be good to your close cultural relatives'; the memetic equivalent of kin selection. But how do you know who they are? This rule tracks biological kinship in cultures with predominantly vertical transmission, since in these cultures you acquire most of your memes from biological relatives, but with horizontal transmission other means of recognition are needed. One is 'Be good to those who act like you'. It works like this. If you see someone else who acts the same way as you do, it is likely that you both have cultural ancestors in common. If you now help him you make it more likely that he will be successful, and hence that he will pass on his memes, including the rule 'Be good to those who act like you'. Allison calls this a 'marker scheme'. He gives the examples of wearing a turban or abstaining from certain foods, but we might add supporting Manchester United or listening to hip-hop, as well as genuflecting or wearing a little portrait of your guru round your neck. He adds that markers that are costly or difficult to learn can deter exploitation by outsiders. Apart from languages, a good example is religious rituals. Many of these require years to learn and others, such as ritual circumcision, are certainly costly for an adult.

The result of this kind of altruism is that people are kind and generous to the in-group and not to outsiders. This boosts the well-being of the group's members and hence makes them more likely to be imitated, and so pass on the faith. This is exactly what we see in many of the world's greatest religions. Although the instruction to 'love thy neighbour as thyself' is commonly taken to mean 'love everyone', in the tribal context in which it was first written it may have been meant more literally - in other words love your own tribe, and your own family, but not everybody else (Hartung 1995). Even the admonition not to kill may originally have applied only to the in-group. Hartung points out that the rabbis of the Talmud used to hold an Israelite guilty of murder if he intentionally killed another Israelite, but killing other people did not count.

Some religions positively encourage murder and war against people of other faiths. Islam has fatwas and jihads to justify killing unbelievers, and especially those who harm or renounce the faith. In February 1989, the Ayatollah Khomeini delivered his famous fatwa on the author Salman Rushdie. This is a direct call to all Muslims to murder Rushdie for daring to blaspheme against the holy Koran in his book Satanic Verses. When the Punishment for renouncing or criticising a religion is so severe, the memes are very ably protected.

Hindus, Muslims, and Christians alike have gone to war again and again in the name of God. When a few hundred Spaniards murdered thousands of Incas, leading to the destruction of an entire civilisation, they did it for the glory of God and the holy Catholic Faith. In a subtler way religious missionaries are still destroying ancient cultures even today. People have been tortured, burned alive, and shot because they believed the wrong thing. Religions teach that God wants you to spread his True understanding to all the world and it is therefore good to maim, rape, pillage, steal, and murder.

We saw how the conspiracy theory protects UFO memes; similar mechanisms protect religious memes. As Dawkins (1993) points out, good Catholics have faith; they do not need proof. Indeed, it is a measure of how spiritual and religious you are that you have faith enough to believe in completely impossible things without asking questions, such as that the wine is really turned into blood. This assertion cannot be tested because the liquid in the cup still tastes, looks and smells like wine - you must just have faith that it is really Christ's blood. If you are tempted by doubt, you must resist. Not only is God invisible but he 'moves in mysterious ways'. The mystery is part of the whole package and to be admired in its own right. This untestability protects the memes from rejection.

Religious memes are stored, and thus given improved longevity, in the great religious texts. The theologian Hugh Pyper (1998) describes the Bible as one of the most successful texts ever produced. 'If "survival of the fittest" has any validity as a slogan, then the Bible seems a fair candidate for the accolade of the fittest of texts' (p. 70). It has been translated into over two thousand languages, exists in many different versions within some of those languages, and even in a country like Japan, where only one or two per cent of the population are Christians, more than a quarter of all households possesses a copy. Pyper argues that Western culture is the Bible's way of making more Bibles. And why is it so successful? Because it alters its environment in a way that increases the chances of its being copied. It does this, for example, by including within itself many instructions to pass it on, and by describing itself as indispensable to the people who read it. It is extremely adaptable, and since much of its content is self-contradictory it can be used to justify more or less any action or moral stance.

When we look at religions from a meme's eye view we can understand why they have been so successful. These religious memes did not set out with an intention to succeed. They were just behaviours, ideas and stories that were copied from one person to another in the long history of human attempts to understand the world. They were successful because they happened to come together into mutually supportive gangs that included all the right tricks to keep them safely stored in millions of brains, books and buildings, and repeatedly passed on to more. They evoked strong emotions and strange experiences. They provided myths to answer real questions and the myths were protected by untestability, threats, and promises. They created and then reduced fear to create compliance, and they used the beauty, truth and altruism tricks to help their spread. That is why they are still with us, and why millions of people's behaviour is routinely controlled by ideas that are either false or completely untestable.

No one designed these great faiths with all their clever tricks. Rather, they evolved gradually by memetic selection. But nowadays people deliberately use memetic tricks to spread religions and make money. Their techniques of memetic engineering are derived from long experience and research, and are similar to those used in propaganda and marketing; with radio, television and the Internet, their memes can spread far further and faster than ever before. Billy Graham's style of tele-evangelism is a good example. He starts by evoking fear, reminding people of all the terrible things happening in the world and of their own impotence and mortality. He presents science as having no answers and as a cause of the world's ills, and then persuades people to surrender to the all-powerful God who is their only hope of salvation. The experience of surrender raises powerful emotions and people turn to God in huge numbers.

Other evangelists use healing to spread the Word. We have seen how perfectly normal psychological processes can make people feel better, even when they are not actually cured, and this is a powerful incentive to take on the God memes that often accompany the healing. The trip to Lourdes is expensive and difficult. Expectations are high. Spiritualist healers are kind and plausible, and really do seem to care about your troubles.

Some use fake healing. In the 1980s, Peter Popoff and his wife Elizabeth brought millions of Americans to God, and millions of dollars to the Popoffs, through their healing missions. Their vast audiences sang and prayed, and watched seriously ill people stagger onto the stage, raising powerful emotions as the Popoffs appealed for donations. As Peter correctly diagnosed illnesses and announced the sufferers cured, people forgot that an hour before Elizabeth had wandered through the audience collecting prayer cards on which people wrote their names, addresses, ailments and other crucial facts. She took these to the computer database backstage and beamed the information to a receiver behind Peter's left ear (Stein 1996).

Miracles of all kinds have been used to convert unbelievers. Jesus walked on water and brought a dead man back to life, nineteenth-century spiritualist mediums created spirit forms made of 'ectoplasm', and the advanced practitioners of transcendental meditation claim to levitate. Some people effectively combine special powers with the altruism trick, such as England's much-loved grandmotherly medium Doris Stokes who packed her audiences with clients whom she already knew, and fooled millions (I. Wilson 1987). Many of those clients were recently bereaved wives, husbands or parents who gained comfort from Stokes's messages but who might have coped better with their grief if they had been helped to accept the reality of death.

I do not mean to imply, from all I have said, that there are no true ideas anywhere in any religion. The memetic mechanisms I have described would allow religions to flourish that were based on complete falsehoods and nothing else, but there may be true ideas embedded in them as well. just as some alternative therapies thrive by including a few treatments that work, so religions may include valid insights as well as misleading myths.

At the heart of many religions lie the mystical traditions, like that found in the fourteenth-century Cloud of Unknowing or the teachings of Julian of Norwich in Christianity; the Sufi teachings of Islam; or the stories of enlightenment in Buddhism. These traditions emphasise direct spiritual experience which is often ineffable and therefore hard to pass on. In spontaneous mystical experiences people typically feel they have been given a glimpse of the world as it really is. They feel that self and other have become one, the entire universe is as it is, or that everything is oneness and light. These may indeed be valid insights (I believe they are), but on their own they are not very successful as memes, and rapidly get overtaken by all the more powerful religious ideas I have described above.

Buddhism provides a good example. If the stories are to be believed, the Buddha sat under a tree, with a fervent desire to understand, until finally he became enlightened. He then taught what he had seen, that everything is empty of self-nature, that life is unsatisfactory, that suffering comes about through craving or attachment, and that the cessation of craving leads to freedom from suffering. He laid down an ethical code of behaviour and taught his disciples to work out their own salvation with diligence, by calming the mind and practising attention in every moment. None of this is very comforting. Basically, it means you are on your own in a fundamentally unsatisfactory world with no one to help you. If you look to anything at all to try to make it better then you are caught up in craving and hence suffering. Enlightenment is not something to be attained; it is simply the giving up of - well everything really. As one of my students put it 'I couldn't bear not to want chocolate. I couldn't even imagine not craving chocolate, let alone not craving anything.' So what happens to difficult ideas like these? Perhaps surprisingly they can and do survive, often by being passed in an unbroken chain from inspiring and enlightened teachers to hard-working pupils. Zen Buddhism sticks quite closely to the simplest teachings and includes no deities or hidden powers; no altruism nor beauty tricks. One is told to find out the truth for oneself and trained simply to sit and watch the mind until it becomes clear. These difficult ideas have survived almost dying out in the East and are now spreading widely in the West (Batchelor 1994). However, other forms of Buddhism are much more popular all over the world, such as Tibetan Buddhism, with its numerous powerful deities, beautiful buildings and paintings, stories of marvellous deeds, reciting of sutras, chants, and liturgies. Whether or not there are true insights at the heart of any religion, the fact is that clever memes will tend to beat them in the battle for replication.

We can now see how and why religions have the power and persistence they do. I want now to consider two further questions. First, have they played any part in meme-gene coevolution? And second, how are religions changing now that memes are being spread by modern technology?

The Coevolution of Religions and Genes

The coevolutionary question is this. Have the religious memes that thrived in the past had any effect on which genes were successful? If so, this would be another example of memetic driving. I shall speculate here and hope that some of the questions I raise may be answered by future research.

We know little of the earliest religions. There is evidence of burial of the dead from the Neanderthals who lived from 130,000 to 40,000 years ago, but it is likely that they were not our ancestors. About 50,000 years ago came what is sometimes called the 'Great Leap Forward', characterised by improvements in toolmaking, the beginnings of art, and the creation of jewellery which was sometimes buried with the dead. We can only guess at religious beliefs but burial rites at least suggest some idea of an afterlife. Modern hunter-gatherer societies have varied religious beliefs, including ancestor worship, special powers attributed to the priest or shaman, and belief in an afterfife. So we might guess that early human religions were something like this.

Early humans lived in bands or tribal societies and only gradually did more complex stratified societies evolve. In chiefdoms or states there is enough division of labour for some people to be completely freed from food production; these are typically rulers of various kinds, together with soldiers and priests. Diamond (1997) argues that the function of ideologies and religions in chiefdoms is to justify the redistribution of wealth, the authority of the rulers, and warfare. Chiefs typically take enormous amounts of wealth from working people and use some of it to build grand temples or public works as visible signs of their power. The people may accept their wealth being taken from them, as they accept taxation in modern societies, if they obtain benefits in return. These benefits may include the reduction of violence within the society, protection from enemies, or facilities for public use. Sometimes the ruler and priest are the same person, but in larger societies separate priests take on the religious functions. The priests promote and police the religious beliefs; the beliefs are then used to justify the conquest of other peoples from whom more goods and power can be stolen.

In memetic terms what this amounts to is that the religious memes are more likely to survive and replicate than competing memes are. For example, religions that required no priests, that took no taxes, or that built no impressive buildings, would have been at a disadvantage. This meant the proliferation of highly organised and stratified societies and of priests who taught and maintained the religion. Religious memes have therefore played an important role in the development of human societies.

The coevolutionary question is whether they have affected the genes along the way. E. O. Wilson (1978) treated religions as a challenge to his new science of sociobiology and speculated about the ways in which religious belief could provide a genetic advantage. For example, religions often include prohibitions against eating potentially contaminated foods, and against incest and other risky sexual activities, and encourage believers to have large and well-protected families. In these and other ways religious belief would benefit the genes of believers and so be expected to continue. The evolutionary psychologist Steven Pinker (1997) has argued that religious beliefs are by-products of the brain modules that were designed to do other things; spirits and gods are based on our concepts of animals and people; supernatural powers are inferred from natural powers; the idea of other worlds is based on dreams and trances. As he puts it: 'religious beliefs are notable for their lack of imagination (God is a jealous man; heaven and hell are places; souls are people who have sprouted wings)' (Pinker 1997, p. 557). These authors argue either that religions provide a genetic advantage, or that they are the by-product of things that once provided genetic advantage. They do not consider the possibility of memetic advantage, nor of memes driving genes.

There are several ways in which memes might have influenced genes. Priests attain power and status by predicting (or appearing to predict) weather, disease, or crop failures; by building or being associated with temples and other grand buildings; by wearing expensive and impressive clothes; and by claiming supernatural powers. In many cultures the priests or rulers are given divine status. We know that women prefer to mate with high-status men, and that these men leave more offspring, either by having more wives or by fathering children by women who are not their wives. Even in societies in which the priesthood is celibate and could not (or at least should not) pass on their genes, other people could acquire power by association. If this religious behaviour helped people acquire more mates, then any genes that inclined them to be more religious in the first place would also flourish. In this way genes for religious behaviour would increase because of religious memes.

The idea of 'genes for religious behaviour' is not at all implausible - all it means is genes that make people more inclined towards religious beliefs and behaviour. Brain development is under genetic control and it is known that some brains are more prone to religious belief and experience than others. For example, people with unstable temporal lobes are more likely to report mystical, psychic and religious experiences, and to believe in supernatural powers, than those with stable temporal lobes (Persinger 1983). Like many other psychological variables, religiosity is known to have a heritable component even today. For example, identical twins are more similar in religiosity than non-identical twins or siblings. In our past there may have been as much genetically controlled variation in religious behaviour as there is now, or even more. If so, two effects are possible. First, the memetic environment could have influenced whether genes for religious behaviour were positively selected or not (increasing or decreasing religious behaviour in general). Second, the religion of the time could have influenced the kinds of genes that survived (i.e. those that produced the kind of religious behaviour best suited to that religion). That would be memetic driving at work.

Group Selection

There is another way in which religious memes might conceivably drive the genes: through group selection. The whole concept of group selection has had a troubled history and been beset by controversy. Earlier this century it was invoked to explain all kinds of behaviours that might conceivably benefit groups or societies, and biologists often appealed to 'group adaptations' or 'the good of the species' without any idea of possible mechanisms. Williams's classic book Adaptation and Natural Selection (1966) pointed out the errors: for example, that selfish individuals could always infiltrate altruistic groups and thrive at their expense. Also groups have a slow lifecycle compared with individuals, and individuals can often move between groups. This means that individual adaptations will almost always predominate over adaptations for the group. Therefore, we should not look to group selection as a force that can make individuals sacrifice their own genetic interests 'for the good of the group'.

Most biologists now consider that group selection is only a weak force in nature (Mark Ridley 1996). However, selection at the level of the group can sometimes occur. Dawkins's distinction between the replicator and the vehicle is helpful here. In most of biology the replicator (the thing that gets copied) is the gene, while the vehicle is the whole organism. Whole organisms; that is individual cats, donkeys, orchids or cockroaches, live or die, and in the process either pass on their genes or not. All the genes in that vehicle share the same fate. In this (the most common) case selection is taking place at the level of the organism.

In some cases, however, whole groups of organisms live or die, and so all the genes in the group are killed off at once. If this occurs then the group is the vehicle and we can say that selection is happening at the level of the group. This applies, for example, to whole species that go extinct, or to isolated populations of animals, such as those on small islands, in which some groups survive and some do not. In these cases there is no conflict between individual and group selection (as there was in the argument about altruistic behaviour) but selection has acted at the level of the group.

Ridley (1996) concludes that group selection works only if migration rates are implausibly low and group extinction rates implausibly high. Another way of putting it is that group selection is favoured by mechanisms that reduce the differences in biological fitness within the groups and increase differences between groups, thus concentrating selection at the group level (D. S. Wilson and Sober 1994).

Memes may provide just this kind of mechanism. Indeed Boyd and Richerson (1990) have used mathematical models to show that group selection is particularly likely to occur when behavioural variation is culturally acquired, and that it can even occur with large groups and substantial rates of migration. The important point is that memes can have precisely the effect of decreasing within-group differences and increasing between-group differences.

Let us take dietary habits as an example. Suppose that one group of people eat shellfish as a major part of their diet and develop ways of cooking mussels or clams and getting them out of their shells, while another group of people hold a taboo against eating shellfish. People within each group are more similar to each other, and different from people in the other group. Migration between groups is made difficult by long habits of taste, and the difficulty of learning how to prepare the food. In some environments the first group may do better because they get more protein, while in other environments the second group may do better because they survive a lethal disease from infected food. When disease strikes or famine threatens it is whole groups that live or die. Food taboos are an important part of many religions. Orthodox Jews do not eat shellfish or pork and avoid mixing meat with milk. Many Buddhists and Hindus are vegetarians because they do not want to kill animals. The beliefs that underpinned these taboos may have caused some groups of people to survive and others to go extinct; and both their genes and their memes would have gone with them.

Religions also dictate sexual practices, promote certain kinds of cooperative behaviour, and regulate aggression and violence. Although many people believe that primitive tribes live an idyllic and peaceful existence, this myth (like so many in anthropology) has been exploded. The anthropologist Napoleon Chagnon (1992) lived for many years with the Yanomamö, who live in the Brazilian rain forest by hunting and growing food in temporary gardens. He describes a violent life in which war between villages is common and murders are revenged with more murders. Similar stories come from many parts of the world. In New Guinea, a group of nomads called the Fayu live in small family groups who only rarely meet other families because of the revenge murders that ensue when they do. Gatherings, for example to exchange brides, are fraught with danger. In many tribal societies murder is a leading cause of death (Diamond 1997). Although many people in modern cities believe that they face ever increasing risks of being killed they are in fact far safer than they would have been in a band or tribal society. The organisation that comes with government and religion therefore decreases these kinds of violence. However, it also provides the justification for large-scale wars.

The history of warfare is largely a history of people killing each other for religious reasons. Religions give people a motive, other than genetic self-interest, for sacrificing their lives for others - something that does not happen in band and tribal societies. Young men may believe that it is good to die for God, heroic to be killed in a religious war, or that they will have a place reserved for them in heaven. A society in which brave young men are prepared to die for their beliefs is likely to win a war against a society in which they are more concerned about protecting themselves or avenging their family. Such a victory is a victory for the memes that created the difference in the first place, and for the genes of the survivors.

We can now see why group selection might be important in memetics. Religions are a good example of a mechanism that decreases withingroup differences, while increasing between-group differences and rates of group extinction. In many religions conformity is encouraged, forbidden behaviours are punished, differences between believers and unbelievers are exaggerated, fear or hatred of people with other beliefs is nurtured, and migration to a different religion made difficult or impossible. Wars between religious groups are common and in our evolutionary history many groups have lived or died for their religion. All this makes it more likely that group selection has occurred. If there were genetic differences between the groups to start with, then the survival of some groups and extinction of others would have had effects on the gene pool. In this case we could say that the religious memes have driven the genes.

This is likely to be most interesting if, for example, there was some genetic reason why one group took up one religion while a different group took up a different religion. Let us imagine two neighbouring groups of early hominids in which, by chance, one group had more of a genetic tendency to want to bury their dead in elaborate ways. This is not at all far-fetched if you remember that digging and burying behaviour is under genetic control in many species, from worms and wasps to rabbits and dogs. This genetic propensity then encouraged these people to develop a religion based on ancestor worship and an afterlife - we can call them the 'Afterlifers'. Meanwhile, the other group developed a religion based on worshipping nature spirits - we can call them the 'Naturists'. The Afterlifers then developed a taste for war, believing their ancestors' spirits would aid them, and that they would individually go to heaven if they killed an enemy, whereas the Naturists just got on with their own interests. In consequence, the Afterlifers won more of the wars against the Naturists; their memes spread - and so did their genes. Genes for the original ritual burying behaviour were selected for by group selection driven by memes.

I am not suggesting that this precise series of events has actually happened, but that this general mechanism could have shaped human nature and given us our religious tendencies. The principle is a general one and could theoretically apply to all kinds of genetic predispositions, such as conformity, having religious experiences, enjoying ritual and worship, or believing in life after death. This process could even have acted to favour genes that would otherwise be detrimental to fitness, or to wipe out genes that would otherwise have been fitness-enhancing. So some aspects of human nature could have been determined not for the sake of the genes but for the sake of the memes. Our beliefs could have moulded the way genetic selection took place. If this has happened it means that human beings might now be naturally religious creatures because of our long memetic history.

Religions have held enormous power for millennia, but times are changing and religions with them. One obvious change is that vertical transmission is giving way to the faster horizontal transmission. As people are increasingly exposed to new ideas from television, radio, newspapers and the Internet, they begin to make comparisons and ask difficult questions. So it is, sadly, not surprising to learn that Afghanistan's Taliban Islamic movement has forbidden televisions and radios, and has set about destroying any they find, and punishing their owners. Meanwhile, in countries with thriving communications, some of the tricks the old religions use may not work so well any more. When people can see films, go to art galleries, and listen to any music they like, the beauty trick is less effective. When we are subjected on television to the gruesome results of religious wars, the altruism trick wears thin. When Christian leaders argue over whether homosexuality is really a sin, the truth trick begins to weaken its grip.

In the past, religions that promoted large families were successful because they created more people to adopt the faith from their parents. Lynch (1996) has given many examples of religions, from the ancient Islam to the relatively new and thriving Mormonism, that spread by increasing the number of their offspring, but he does not clearly differentiate the effects of vertical and horizontal transmission. With modern horizontal transmission people are less bound by their parents' beliefs; as memes spread faster and faster the birth rate becomes less and less significant. We should therefore expect proselytic religions to do better in technologically advanced societies. We may expect new religions of this kind, and also that old faiths which can adapt their memes to changing times may survive while others will become extinct.

I doubt that human beings will ever be entirely free of religion. If the arguments above are right then religions have two very strong forces going for them. First, human minds and brains have been moulded to be especially receptive to religious ideas, and second, religious memes can use all the best meme tricks in the book to ensure their own survival and reproduction. This may explain the persistence of religion in scientifically literate societies and in societies in which political dogma has tried to erase all religious behaviour - and failed. Perhaps our brains and minds have been moulded to be naturally religious and it really is difficult to use logic and scientific evidence to change the way we think - difficult, but not impossible.

Science and Religion

I have implied that science is, in some sense, superior to religion, and I want to defend that view. Science, like religion, is a mass of memeplexes. There are theories and hypotheses, methodologies and experimental paradigms, intellectual traditions and long-standing false dichotomies. Science is full of ideas that are human inventions, and have arbitrary conventions and historical quirks built into them. Science is not 'The Ultimate Truth' any more than any other memeplex. However, memetics can provide a context in which to see why science offers a better kind of truth than religion.

We are designed by natural selection to be truth-seeking creatures. Our perceptual systems have evolved to build adequate models of the world and predict accurately what will happen next. Our brains are designed to solve problems effectively and to make sound decisions. Of course, our perception is partial and our decision-making less than brilliant - but it is a lot better than useless. If we had no memes, that would be that; we would have the best understanding of the world that could be acquired in the circumstances. But we do have memes, and with memes come not only new ways of controlling and predicting the world, but meme tricks and free-loading memes, misleading memes and false memes.

Science is fundamentally a process; a set of methods for trying to distinguish true memes from false ones. At its heart lies the idea of building theories about the world and testing them, rather like perceptual systems do. Science is not perfect. Scientists occasionally cheat to gain power and influence, and their false results can survive for decades, misleading scores of future scientists. False theories thrive within science as well as within religion, and for many of the same reasons. Comforting ideas are more likely to last than scary ones; ideas that exalt human beings are more popular than those that do not. Evolutionary theory faced enormous opposition because it provided a view of humans that many humans do not like. The same will probably be true of memetics.

However, at the heart of science lies the method of demanding tests of any idea. Scientists must predict what will happen if a particular theory is valid and then find out if it is so. That is precisely what I have tried to do with the theory of memetics.

This is not what religions do. Religions build theories about the world and then prevent them from being tested. Religions provide nice, appealing and comforting ideas, and cloak them in a mask of 'truth, beauty, and goodness'. The theories can then thrive in spite of being untrue, ugly, or cruel.

In the end, there is no ultimate truth to be found and locked up forever, but there are more or less truthful theories and better or worse predictions. I do defend the idea that science, at its best, is more truthful than religion.

Susan Blackmore is a Senior Lecturer in Psychology at the University of the West of England, Bristol, where she lectures on the psychology of consciousness. She is the current Perrot-Warrick Researcher, studying psychic phenomena in borderline states of consciousness, and has recieved the Distinguished Skeptic's Award from CSICOP. Her official website can be found at http://www.uwe.ac.uk/fas/staff/sb/index.htm.