Also, genes are not the only other replicators to consider. For example, our immune systems are now known to work by selection. The British psychologist Henry Plotkin (1993) refers to both brains and immune systems as ‘Darwin machines’, and in his study of Universal Darwinism uses general evolutionary theory to apply to many other systems including the evolution of science. In each case, one can apply the ideas of replicators and vehicles (or of replicators, interactors and lineages, to use Hull’s formulation) to understand the way the system evolves.
We should think of it like this – evolutionary theory describes how design is created by the competition between replicators. Genes are one example of a replicator and memes another. The general theory of evolution must apply to both of them, but the specific details of how each replicator works may be quite different.
This relationship was clearly seen by the American psychologist Donald Campbell (1960, 1965) long before the idea of memes was invented. He argued that organic evolution, creative thought and cultural evolution resemble each other and they do so because all are evolving systems where there is blind variation among the replicated units and selective retention of some variants at the expense of others. Most importantly, he explained that the analogy with cultural accumulations is not from organic evolution
per se
; but rather from a general model of evolutionary change for which organic evolution is but one instance. Durham (1991) calls this principle ‘Campbell’s Rule’.
We need to remember Campbell’s Rule when we compare memes and genes. Genes are instructions for making proteins, stored in the cells of the body and passed on in reproduction. Their competition drives the evolution of the biological world. Memes are instructions for carrying out behaviour, stored in brains (or other objects) and passed on by imitation. Their competition drives the evolution of the mind. Both genes and memes are replicators and must obey the general principles of evolutionary theory and in that sense are the same. Beyond that they may be, and indeed are, very different – they are related only by analogy.
Some critics have tried to dismiss the whole idea of memetics on the grounds that memes are not like genes, or that the whole idea of memes is only an ‘empty analogy’. We can now see why these criticisms are misguided. For example, Mary Midgley (1994) calls memes ‘mythical entities’ that cannot have interests of their own, ‘an empty and misleading metaphor’, a ‘useless and essentially superstitious notion’. But Midgley has misunderstood the way in which replicators can be said to have power or ‘interests of their own’ and therefore she simply misses the strength and generality of evolutionary theory. Memes are no more ‘mythical entities’ than genes are – genes are instructions encoded in molecules of DNA – memes are instructions embedded in human brains, or in artefacts such as books, pictures, bridges or steam trains.
In a radio debate, Stephen Jay Gould (1996
b
) called the idea of memes a ‘meaningless metaphor’ (though I am not sure one can actually have a meaningless metaphor!). He goes even further and rejects the very notion that ideas and culture can evolve, pleading ‘I do wish that the term
“cultural evolution” would drop from use’ (Gould 1996a, pp. 219-20), but I do not think that it will, because culture does evolve.
Gould seems to think that because memes and genes are related by analogy or metaphor we would somehow be doing a disservice to biological evolution by making the comparison. Again, he has missed the point that both are replicators but they need not work in the same way.
My own view is that the idea of memes is an example of the best use of analogy in science. That is, a powerful mechanism in one domain is seen to operate in a slightly different way in an entirely new domain. What begins as an analogy ends up as a powerful new explanatory principle. In this case, the most powerful idea in all of science – the explanation of biological diversity by the simple process of natural selection – becomes the explanation of mental and cultural diversity by the simple process of memetic selection. The overarching theory of evolution provides a framework for both.
With Campbell’s Rule in mind we can now go on to the task of trying to understand the evolution of memes. We may use the gene as an analogy but must not expect too close a comparison. Instead, we must rely on the fundamental principles of evolutionary theory to guide us in understanding just how memes work.
Copy me!
What is special about the sentence ‘Say me!’ – or ‘Copy me!’ – or ‘Repeat me!’?
They are simple (perhaps the simplest possible) examples of self–replicating sentences. Their whole point is to get themselves copied. These sentences are certainly memes – but probably not very effective ones. I doubt you will now go around shouting ‘Say me!’ to all your friends, but there are tricks that can be added to the simpler sentence to improve its copying potential. Hofstadter (1985) wrote about such ‘viral sentences’ in his monthly column in the magazine
Scientific American
called ‘Meta–magical Themas’, and readers wrote in with many more examples.
Take: ‘If you copy me, I’ll grant you three wishes!’ or ‘Say me or I’ll put a curse on you!’ Neither of these is likely to be able to keep its word and few people over the age of five are likely to fall for such simple–minded threats and promises. Unless – Hofstadter adds – you simply tack on the phrase ‘in the afterlife’.
In fact, it is often at about the age of five that many of us meet such sentences for the first time. I well remember being excited when I received
in the post a letter that contained a list of six names and instructed me to send a postcard to the first name on the list. I was to put my own name and address at the bottom and send the new list to six more people. It promised me I would receive lots of postcards. I do not remember whether my mum prevented me from joining in or not. She might have been wise to – realising, though she would not have put it that way, that my meme–immunological system was not yet well developed. I certainly do not remember a deluge of postcards.
As these things go, that was a fairly innocuous chain letter, consisting of just a promise (the postcards) and an instruction to pass it on. At worst I would have wasted seven stamps and a postcard. I might even have received a few cards myself. Many are much more sinister, such as pyramid selling schemes, that can lose people fortunes. You would think such trivial schemes would die out, but they do not seem to. Only recently, I received an e–mail that said ‘Do you like to play those scratcher lottery tickets?’ (I do not) ‘Would you like to learn how to turn 6 tickets into thousands?’ (not particularly) ‘You’ll receive lottery tickets from all over the country every month! Have fun just collecting them or scratch them for the $$$jackpot$$$. There is a free service on the Web that can set you up to do just that!’ Do people really join up? I suppose they must.
These are all examples of groups of memes that are replicated together. Dawkins calls such groups ‘coadapted meme complexes’, a phrase recently abbreviated to ‘memeplexes’ (Speel 1995). Memetic jargon is changing so fast and much of it is so poorly thought out and so misused that I shall try to avoid using it. However, ‘memeplex’ is a handy word for an important concept and so it is one of the few new words I shall adopt.
Genes, of course, go around in groups too. They clump together into chromosomes, and chromosomes are packed together inside cells. Perhaps more importantly, the whole gene pool of a species can be seen as a group of mutually cooperating genes. The reason is simple: a free–floating piece of DNA could not effectively get itself replicated. After billions of years of biological evolution, most of the DNA on the planet is very well packaged indeed, as genes inside organisms that are their survival machines. Of course, there are occasional ‘jumping genes’ and ‘outlaw genes’ and little bits of selfish DNA hitchhiking on the rest, and there are viruses that are minimal groups that exploit the replicating machinery of other larger groups – but groups, by and large, are necessary for genes to get around at all.
We could simply draw the analogy and say that memes should behave the same way but it is better to go back to the basics of evolutionary
theory. Imagine two memes, one ‘send a scratchcard to
x’
and another ‘win lots of money’. The former instruction is unlikely to be obeyed just on its own. The latter is tempting but includes no instruction on how to. Together, and with some other suitable co–memes, the two can apparently get people to obey – and copy the whole package on again. The essence of any memeplex is that the memes inside it can replicate better as part of the group than they can on their own. We shall meet many more examples of memeplexes in due course.
The simple self–replicating meme groups we have considered so far have been given a great boost by the advent of computers and the Internet. Computer viruses are an obvious and familiar example. They can leap from user to user and the number of users (at least at the moment) keeps increasing. They can cross vast distances at the speed of light and then lie dormant in safe and solid memory banks. However, they cannot be just a bare instruction to ‘Copy me’. This might succeed in clogging up the entire memory of the first computer it got into but would have no way of getting any further. So viruses have co–memes for promoting their survival. They lurk in the programs that people mail to their friends on disks. Some evade immediate detection by infecting only a small proportion of the machines they reach, and some are triggered probabilistically. Some bury themselves in memory only to pop up at a specified time – we may expect many at midnight on 31 December 1999 – quite apart from the looming problem of computers that cannot cope with the year ‘00’.
Some have quite funny effects, such as making all the letters on a computer screen fall to the bottom of the page – with a devastating effect on the user, but some have clogged up entire networks and destroyed books and doctoral theses. My students have recently encountered a virus on the word processor Word 6.0 that lives in a formatting section called ‘Thesis’ – tempting you to get infected just when your year’s work is almost finished. No wonder networks are now protected by frequent automated virus checks and we have a proliferation of anti–virus software – medication for the infosphere.
Internet viruses are a relatively new arrival. I once received ‘Penpal Greetings’, apparently a very kind warning from someone I have never met. ‘Do not download any message entitled “Penpal Greetings” ’ it said –and went on to warn me that if I read this terrible message I would have let in a ‘Trojan Horse’ virus that would destroy everything on my hard drive and then send itself on to every e–mail address in my mail box. To protect all my friends, and the worldwide computer network, I had to act fast and send the warning on to them.
Have you spotted it? The virus described does not make sense – and does not exist. The real virus is the warning. This is a very clever little memeplex that uses both threats and appeals to altruism to get you – the silly, caring victim – to pass it on. It is not the first – ‘Good Times’ and ‘Deeyenda Maddick’ used a similar trick. ‘Join the Crew’ is slightly more damaging, warning ‘do not open or look at any mail that says
RETURNED OR UNABLE TO DELIVER
. This virus will attach itself to your computer components and render them useless. Immediately delete … there is NO remedy’. Anyone who does not spot the trick will presumably delete all those messages they had sent to people whose addresses have changed or whose e–mail systems were temporarily unavailable. A little bit of self–replicating code, using a combination of humans and computers as its replicating machinery, can have annoying consequences.
What will happen next? As people become familiar with these viruses they may learn to ignore the warnings. Thus the original type of virus will start to fail but it might let in something worse, as people start ignoring warnings they ought to heed. But then again, if ordinary old–fashioned chain letters still work, perhaps things will not change so very rapidly.
All this talk of viruses makes me wonder just why we call some pieces of computer code a virus and others a computer program. Intrinsically, they are both just lines of code, bits of information or instructions. The word is, of course, taken directly by analogy from biological viruses and probably based on the same intuitions about the way these bits of code spread. The answer is not so much to do with the harm they do — indeed some really do very little – but to do with their function. They have none apart from their own replication.
Bacteria are more complex than viruses and can be positively helpful as well as harmful. Lots live in symbiosis with us and with other animals and plants. Many do important jobs inside our bodies. Some have been co–opted to make special foods for us and so on. Viruses do little else than replicate themselves – and then only by stealing other organisms’ replicating machinery. So the comparison with today’s rather simple computer viruses is apt.
Might we build the equivalent of computer bacteria? Perhaps this would be a better term for some existing programs that are deliberately used to infect computer systems and run around doing jobs like updating databases or seeking out errors. Dawkins (1993) imagines useful self–replicating programs that might carry out market research by infecting many computers and then, as occasional copies got back to their starting place, providing useful statistics on user habits. Simple robotic programs, or bots, are already designed to roam around congested communications
networks leaving trails that provide information about the best and worst areas of congestion, or to mimic human users in games and virtual environments. Might such simple creatures gang up together to create powerful groups just as genes have done?
These ideas seem to stretch the analogy with biological viruses a bit far (and we must be very careful of such analogies), but they do remind us that replicators vary in their usefulness. We tend to call something a virus when it is clearly acting mainly for its own replication by stealing the replicating resources of some other system – and especially when it does harm to that system. We usually give it a different name when it is useful to us.