Memetics

Memetics
Full disclosure: I think that natural selection is a force of nature that is on a par with gravity in significance. It was discovered by Darwin in the context of biology but (for me) the idea was generalized by Richard Dawkins and Daniel Dennett into the Evolutionary Algorithm. Wherever that algorithm (EA) can work it does work. Dennett called the EA "substrate neutral". Arithmetic is a substrate neutral set of algorithms. It doesn't matter if you execute them with pen and paper or with a calculator or in your imagination.

Darwin first articulated the algorithm of natural selection. The algorithm itself is pretty simple though it's implications are huge. For completeness I'll spell out a simple version.
First he observed that parents had children. The children inherited traits from the parents but were never exactly the same as parents. Among the children there would be some whose traits enabled them to have more children themselves.

This had been known since ancient times. After all, this is the basis of animal and plant breeding. The breeder chooses what to breed and what to eat. After a while you have a new breed.

What Darwin realized was that in the natural world parents produce variable children and the ecosystem itself provides a selection pressure that determines which sort of child populates the future. He looked at a grassy bank with all it's plants and animals and realized that they were all providing a selection pressure on each other.

I'd always sort of understood Natural Selection - but the version I got in high school wasn't really the case. Nature doesn't care for anything but it's not all "red of tooth and claw" either. There is a lot of symbiosis and mutualism and co-operation too. But it wasn't until I encountered Daniel Dennett's "Consciousness Explained" that I fully got the power and beauty of Natural Selection.

One of the great gifts I got from that book was his generalization of Natural Selection into the Evolutionary Algorithm which goes like this:
Do in a loop
A replicator makes a replicant
The replicants have small variations
Some of those variations enable the replicant to become a more prolific replicator
Repeat loop

What Darwin knew was that he could create a distinct breed of pigeon with a small number of generations. What he realized was that over millions of generations we could get all of the wonder of the natural world from very simple beginnings. But for Darwin the replicator was the parent and the replicant was the child. It turned out that that's not quite the case.

Gregor Mendel studied peas and found that characteristics like spotted or wrinkled or not seemed to pass to children independently of each other. He speculated that these characteristics were caused by invisible internal "factors" that later were called genes.
By the 1950s we actually understood the genetic mechanism at a molecular level. Individual genes were observed to be transmitted without change through many generations.

In the mid 1970s, Richard Dawkins (a biologist at Oxford) realized that it was the genes in DNA that were the actual replicators. For instance, if a dog has puppies the dog isn't the replicator. The genes she passes to the puppies are the replicator.

Each of us has the same DNA in the nucleus of each of our cells. That's our genotype. In a marvelous developmental process that genotype kind of "unfolds" to produce a phenotype - a physical body with various characteristics.

So the replicator (genes in DNA) can only do that if they produce a phenotype that is successful in the world and survives and breeds.

Dawkins likens this to the phenotype being like some sort of autonomous robot that carries the genotype from generation to generation. (Aside: Keith Stanovich in "The Robot's Rebellion" explores the possibility that humans may be able to rebel against their genome's interest).

Once it is understood that Natural Selection is working at the genomic molecular level rather than at the level of phenotype, that Natural Selection was a mechanical process that didn't depend on anything doing a mental process of 'selecting' we are at the Evolutionary Algorithm.

(Note: I discovered Dawkins via Dennett and slurped his ideas right up. I do remember hearing of the Selfish Gene when it first came out and sniffing "that's crazy - genes can't be selfish". Turns out that selfish doesn't quite mean what we thought :-)

Famously, Dawkins asked himself whether there were other situations where the EA might apply and (I think to his surprise) found it. He observed that people imitate each other's behavior and that the imitation can vary and some imitated behavior gets imitated more often than others. This looked a lot like the EA in operation. He coined the term 'meme' for that sort of imitated behavior.

At first glance it seems that imitated behaviors do evolve over time as they get imitated again and again. And I think over the decades the demonstration has been done again and again. To me the whole marketing industry is practical memetics though they don't use that terminology.

With memetics we need to hold the idea that memes are imitated behaviours close to our hearts. The idea of 'meme' is itself a meme and its various species have evolved away from the core idea and a problem emerged because the meaning of meme evolved and not all of those ideas work very well.
There is a process similar to speciation that happened.
For instance one thread saw memes as 'units of culture' and I never thought that worked very well.
There is a category problem: a culture is not really in the same category as a phenotype and memes don't seem to relate to culture the same way that genes relate to phenotypes at all.
Mary Midgley puts it like this ". . . .culture is not best understood by examining its smallest parts, as culture is pattern-like, comparable to an ocean current. Many more factors, historical and others, should be taken into account than only whatever particle culture is built from. "

Another species of 'meme' (which I quite like) involves an image with some sort of short meaningful text superimposed in large bold outlined text.

Another species is one that poo poos the whole idea of meme as replicator that evolves by saying that memes are nothing like DNA. After all, DNA is a long sequence of a particular set of molecules linked together in a way that can be read and interpreted. Memes offer nothing like that. Which is true. Not a problem. Memes are imitated behaviours - they are not molecules.

Memes and genes share certain properties because they each evolve according to the Evolutionary Algorithm. Things that evolve that way happen in populations. Genes are found accompanied by many other associated genes. A gene floating in water without that community is pretty meaningless. But also within DNA there are groups of genes that remain together through the generations. For one reason or another they are mutually reinforcing.

That idea of different replicators who enhance each others' reproductive success by the happenstance of being together. Sometimes the relation is loose, like in ecosystems. Sometimes it's tight as with symbionts. I think it's a natural tendency in evolving systems and in biology we see it working on every level from the molecular on up.

Memes don't exist alone. They are always associated with many other memes in a population and in individual structures of understanding. The same thing happens. Some memes mutually reinforce each other. The effect of the mutual reinforcement can be strong enough that groups of memes can become bound into structures called a memeplex. Things like science or religion are memeplexes. So is the State. But memeplexes don't have to be so grand. Conspiracy theories present a less grandiose example. Within memeplexes you can even find memes that defend the whole. For example, if you say to a conspiracy theorist that you see no evidence of the conspiracy, the defensive meme is deployed: "See???? That proves how strong the conspiracy is!!!"

Memetics depend on people with all of their fabulous cognitive capacities as the substrate just as genetics depends on the fabulous capacities of chemistry. Since memes exist on a substrate of many interacting minds they are independent of any particular mind.

Here we run into a kind of language problem: we use various descriptive metaphors for this stuff that is not strictly speaking accurate. We speak of memes jumping from mind to mind and we speak of that jumping often as an infection. These ideas speak to the fact that the memes do that jumping because they can not because anybody (not even the meme) thinks its a good. A meme isn't really a 'thing' that can do anything. It's an imitated behavior.

Memes are transmitted in many ways. Parents are passing on memes when they teach kids to speak. The kids unconsciously pick up memes when they take on the parents accent. Teachers pass on memes. Advertisers know all about memes and how to pass them on subliminally (think glamorous people in glamorous settings smoking in movies long after smoking was known to be dangerous. Memetics? You decide :-)

I admit that unless I'm doing philosophy memetics rarely appears in my daily life. When I go out every day to get groceries I'm not imitating a behavior even though I'm repeating a behavior. When I rode my bike across Canada, the idea of the trip may have been a meme, but the day to day experience sure wasn't. I understand that when something is explained in terms of memetics there are usually other ways of explaining the same thing using other explanatory structures.

So what's so good about memetics?

First up - personally I like the concept. It fits snuggly within my own structure of understanding. Associated with that; I've been working with the idea for decades now and have found it very useful in trying to understand what I see happening in society. That aid to understanding is important to me but not to everyone.

I've found that memetics is evidence for the Evolutionary Algorithm. Darwin thought that natural selection took place across thousands or millions of generations and the rate of change would be too slow for humans to see directly. (In fact it's been shown in some cases how evolution can be detected within the human attention span). Memes evolve very much faster than that. And once you know where to look, Memes are all over the place and easy to see.

And note: I say evolve rather than "evolve" or "seems to evolve". The contention there is that Natural Selection is contained within the set of algorithms I call Evolutionary Algorithms. Other EA's are not simulations of Natural Selection. They are instead other instances of the EA.

A big part of my own philosophic stance is the need to 'ground' abstractions like 'cognition' in physical reality. Memetics does that. Fashion is a good example. When I walk the streets there are discernable styles that people are wearing. I know that this is no accident. It takes a long time and a lot of work to make clothes and bring them to market. Stylemakers don't do that as a gamble. They use focus groups to get a sense of what sort of meme will be popular. They associate the products with various high status people knowing that people imitate the behavior of high status people. I think memetics explains why that works. And for me it's a link in the process of grounding that sort of thing (fashion) in physical reality.

As always, the level of explanation provided by memetics is very far from being a complete explanation. For instance, we've talked about the power relationships implied by wage labor. Memetics does not replace such an explanation at all. At the same time, power relationship explanations don't replace memetics either. This is a normal thing. After all, physical explanations for events don't replace memetic explanations or power relationship explanations. They are all there at once. It's a matter of context - you use the explanation that is relevant to the context.

Martin Hunt
2021