Sunday, 8 October 2017

Minsky on a new kind of evolution

Here's the late Marvin Minsky (1994) in "Will Robots Inherit the Earth?":
In the past, we have tended to see ourselves as a final product of evolution - but our evolution has not ceased. Indeed, we are now evolving more rapidly - although not in the familiar, slow Darwinian way. It is time that we started to think about our new emerging identities. We now can design systems based on new kinds of "unnatural selection" that can exploit explicit plans and goals, and can also exploit the inheritance of acquired characteristics. It took a century for evolutionists to train themselves to avoid such ideas - biologists call them 'teleological' and Lamarckian' - but now we may have to change those rules!
That's more or less what I have been trying to do over the last decade: drag the theory of evolution into the 21st century by incorporating intelligent design, Lamarckain inheritance, directed mutations, evaluation under simulation and so on.

One of the things I have found is that these things are often not quite as novel an Minsky implies. Organisms have been "inheriting acquired characteristics" for at least as long as dogs have been passing their fleas on to their puppies. Plans and goals are not exactly new either. The first mammals were making plans - and these went on to influence their evolution via sexual selection and in other ways. The picture of these new capabilities arising with human engineering design is not really correct - many of them have much older roots.

IMO, this is interesting because it makes the old school evolutionary biologists and their textbooks wrong in their own terms, not just because of human beings, genetic enginnering, etc.

Sunday, 24 September 2017

Megaclones

Large-scale cloning is common in both organic and cultural evolution. Multi-cellular organisms are largely clones of a single genotype, though some "somatic mosaicism" does happen. In the cultural realm, there are large-scale clones of a variety of books, music files, vidoes and pieces of computer software.

Though there's not much difference between organic and cultural evolution in this respect, they do seem a little bit different when it comes to spatially-distributed megaclones. In the organic realm, megaclones are mostly single organisms. Asexual reproduction does produce a similar effect. For example, dandelions reproduce asexually, and nearby dandelion plants are often closely related. However, there's no coordination maintaining the genetic similarity, and so over time the genomes diverge.

In the cultural realm, if you look at software like Android or iOS, these are massive distributed megaclones. These are good examples of cultural eusociality. The manufacturer is like the queen, while the individual phone handsets are like drones. Unlike the situation with ants or bees, variation due to sexual recombination is pretty minimal - so the whole system is closely related and can be modelled as being a single distributed cultural organism.

Money is another classic example of a large-scale distributed cultural megaclone. The notes and coins are typically identical on a large scale (not counting their serial numbers). Here the "queen" is the mint, while the notes are the "drones".

Megaclones are often important determinants of what counts as an evolutionary unit. A megaclone can be modelled as an individual, or an organism, without too much concern for conflict between the cloned units.

That distributed megaclones seem more viable in the cultural realm has an important effect on the evolutionary dynamics involved - namely cultural evolution has lifted the size limit on organisms. Blue whales are pretty big animals, but cultural megaclones, can span the entire planet easily these days. It looks as though some future organisms will be enormous.

Sunday, 17 September 2017

Modern anti--Natalism

The demographic transition describes how rich countries often wind up with sub-replacement fertility levels. To sustain their populations each woman needs to have at least two children. Yet in Japan, the fertility rate is 1.4. In South Korea, it is 1.3. In Hong Kong, it is 1.2. In Taiwan it is 1.1. For more stats on this see here.

This is a bit of a puzzle for the "at all boils down to DNA genes" versions of evolutionary theory (i.e. most sociobiology and evolutionary psychology) - since basic theory predicts that the more resources you give an organism, the more offspring they are expected to have.

The standard memetic explanation for this is that memes compete with genes for resources and act to divert host resources from making more DNA genes to making more memes. Dawkins gave essentially this explanation in 1976, referring in particular to the low fertility of priests - and how their resources were being directed away from gene propagation and into meme propagation.

Conventional explanations of the phenomena observe that famale choice is involved. Years of college education in girls is strongly negatively correlated with fertility. Educated girls are waiting longer before having kids and are then having fewer of them. Another fairly obvious factor is cheap family planning technology.

There are some explanations that don't involve memes. For example the r/K selection axis is fairly clearly involved - and it is possible that some environments act as superstimulii for faculative K-selection mechanisms - producing a maladaptive response. Faculative K-selection mechanisms are certainly part of the story, but they are more like one of the targets that the memes use to effect their results than the main story.

I think the main mechanism is that influential female role models tend to be those who have prioritised meme reproduction over their DNA genes. While mothers are busy raising their children they have less time and resources available for influencing others. On the other hand, take Jennifer Aniston, for example. Apparently she has said: "I've never in my life said I didn't want to have children. I did and I do and I will... I would never give up that experience for a career." However, somehow or another she still doesn't have any kids. Women who have prioritized their career over having kids are more likely to be publicly-visible role models - leading to a meme-driven plague of female infertility.

I am fascinated by self-conscious expressions of this tendency. Steven Pinker provided me with some early examples:

Well into my procreating years I am, so far, voluntarily childless, having squandered my biological resources reading and writing, doing research, helping out friends and students, and jogging in circles, ignoring the solemn imperative to spread my genes.

...and...

By Darwinian standards I am a horrible mistake, a pathetic loser, not one iota less than if I were a card-carrying member of Queer Nation. But I am happy to be that way, and if my genes don't like it, they can go jump in the lake.

Dawkins famously wrote:

We are built as gene machines and cultured as meme machines, but we have the power to turn against our own creators. We, alone on earth, can rebel against the tyranny of the selfish replicators.

Keith Stanovich turned this into a manifesto, with his book: The Robot's Rebellion: Finding Meaning in the Age of Darwin.

Evangelical approaches to the issue are of particular interest. For example consider Richard Stallman's essay:

Why it is important not to have children. Richard starts out provocatively with:

The most important thing you can do, to avoid global disaster and make a positive contribution to the world, is avoid having children.

He goes on to explain:

Overpopulation is a tremendous danger to civilization and the ecosphere. It makes every human-caused ecological problem bigger. Population growth has slowed but not stopped. The human population is expected to grow by 2 or 3 billion by 2050, and it is not clear how to find water and food for all those people. Population growth also increases the difficulty of curbing global heating. Thus, the decision about having children is, for most people, the most important decision in their lives about how they will affect humanity's resource footprint in the future.

Others have also put a moral spin on the issue.

David Benatar's "Better Never to Have Been: The Harm of Coming into Existence" argues that having kids is always bad because it introduces more suffering into the world.

Andrés Gómez Emilsson recently wrote:

Your selfish genes will try to do everything they can to make you feel like not reproducing is the same as dying and going to hell. For the love of God, do not listen to your selfish genes.
Dawkins (again) wrote:

"As for me, I'd rather spread memes than genes anyway.

One of the biggest modern anti-Natalist experiments was the Chinese "one child" policy - which made having multiple kids illegal.

I am more pro-natal than anti-Natal. I'm not especially evangelical on the issue, though I do describe some of the anti-natalists as being "pro-death" and generally warn against their influence.

I think Stallman is mistaken in thinking that more people will make the world worse. I am more with Julian Simon in The Ultimate Resource - more people are better. China looks set to be a big force in the 21st century. Their secret is that they have more people - and that means more scientists, engineers and other folk that make the world a better place. Overpopulation seems like a very distant hazard to me - the carrying capacity of the planet is clearly enormous.

Underpopulation is much more serious problem. This century is likely to see "peak human" - as people spend more and more time in computer-generated environments and in the company of sexbots and virtual catwomen. As machines rise, the human gene pool is likely to falter and then fall. Anti-natalism will be part of how it happens. I generally favour slow transitions over fast ones. I don't think procreation will save the humans from being made redundant by technology, but lack of procreation could lead to a more rapid demise for humans, and a rapid transition increases the chance of important information getting lost during the transition.

As for the moral dimension of Darwinism, that debate dates back to Huxley and Kropotkin. Huxley argued that nature was bad:

From the point of view of the moralist the animal world is about on a level of a gladiator’s show. The creatures are fairly well treated, and set to fight – whereby the strongest, the swiftest, and the cunningest live to fight another day. The spectator has no need to turn his thumbs down, as no quarter is given. [...] But, in civilized society, the inevitable result of such obedience [to the law of bloody battle] is the re-establishment, in all its intensity, of that struggle for existence – the war of each against all – the mitigation or abolition of which was the chief end of social organization.

...while by contrast, Kropotkin saw evolution as leading to cooperation and morality. My position is much more on Kropotkin's side than Huxley's. Yes, evolution has produced some suffering, but give it a chance: it hasn't finished booting up yet.

Sunday, 3 September 2017

Max Tegmark's evolutionary classification scheme

I read Max Tegmark's article in Scientific American promoting his book, Life 3.0.

Tegmark proposes the following classification scheme:

In summary, we can divide the development of life into three stages, distinguished by life’s ability to design itself:

  • Life 1.0 (biological stage): evolves its hardware and software
  • Life 2.0 (cultural stage): evolves its hardware, designs much of its software
  • Life 3.0 (technological stage): designs its hardware and software
This isn't a classification scheme I have heard of before. Tegmark introduces the scheme by saying:

I find it helpful to classify life forms into three levels of sophistication: Life 1.0, 2.0 and 3.0.
My first reaction was that these categories were three of the floors in the Tower of Optimization classification scheme I proposed back in 2011.

My second reaction was that Tegmark's numbering scheme seems pseudoscientific. I named my tower floors, rather than numbering them to better allow for future insertions and deletions. However Tegmark only has three categories.

There's an existing literature on the major evolutionary transitions. To say that scientists don't agree with Tegmark's classification scheme seems like a big understatement to me. Numbering schemes seem rather premature.

In my essay, I at least cited some prior work in the field - while Tegmark doesn't seem to have any citations at all. Presumably Max Tegmark made this classification scheme up. It seems like an example of how not to perform scientific classification to me.

The World Made Meme

A new book on Internet memes was published by MIT press in 2016. It is called "The World Made Meme" and it is by Ryan M. Milner. Here is the MIT press page about the book. The book has 272 pages and there are hardback and paperback editions.

I have very briefly skimmed the book in a boostore. It has a large number of pictures of image macros in it, along with a lot of accompaning text. The blurb explains that the book is about internet memes and their effect on public conversations. I'll try to review the book in due course.

This is the second MIT press book on internet memes in recent years. While I look forward to there being more, scientists really need to work on memes more than internet memes. It's true that internet memes are the latest, shiniest type - but it all seems rather like Darwin writing about earthworms rather than evolution.

Sunday, 27 August 2017

The scientific neglect of filtering and sorting

In computer science, filtering and sorting are big topics. Donald Knuth devoted volume 3 of his epic The Art Of Computer Programming to these two topics. That's a reasonable indication of their importance in computer science.

However, science in general took a different route. Filtering is dealt with partly as "selection" - which is covered most comprehensively by evolutionary biology. However filtering is a much broader topic, which extends well beyond biology. As a result, the science of selection is fragmented:

Because the science involved is fragmented there are also a number of areas where it could be applied, but currently isn't - because its influence is not understood or recognised. When small coins accumuate in your wallet, that's a type of selection. Similarly selection results in unpalatable goods accumulating in your refridgerator. Selection is also important in many common physical phenomena, such as erosion, crack propagation, catalysis, crystal growth and electrical discharges. However, its influence often goes unrecognised there as well.

In 1971, George Price called for a theory of selection, writing:

A model that unifies all types of selection (chemical, sociological, genetical, and every other kind of selection) may open the way to develop a general ‘Mathematical Theory of Selection’ analogous to communication theory.

Price continues with:

Selection has been studied mainly in genetics, but of course there is much more to selection than just genetical selection. In psychology, for example, trial-and-error learning is simply learning by selection. In chemistry, selection operates in a recrystallisation under equilibrium conditions, with impure and irregular crystals dissolving and pure, well-formed crystals growing. In palaeontology and archaeology, selection especially favours stones, pottery, and teeth, and greatly increases the frequency of mandibles among the bones of the hominid skeleton. In linguistics, selection unceasingly shapes and reshapes phonetics, grammar, and vocabulary. In history we see political selection in the rise of Macedonia, Rome, and Muscovy. Similarly, economic selection in private enterprise systems causes the rise and fall of firms and products. And science itself is shaped in part by selection, with experimental tests and other criteria selecting among rival hypotheses.

If the situation with filtering in science is bad, the situation with sorting is surely worse. At least selection is championed by evolutionary biologists. Sorting is also very common. You can see its results while looking at stones on a beach or clouds in the sky. Shaking your breakfast cerial makes the biggest lumps rise to the top - a simple sorting operation. I have talked about "natural sorting " before - but most people have never heard of it. If the science of filtering is fragmented, the science of sorting is positively obscure.

There have been efforts to build a general science of selection. Proponents of universal Darwinism have been working on it. There's Zukav's "Without Miracles". There's Hull's "Science and selection". There's Fog's "Towards a universal theory of competition and selection". There's Campbell's "Epistemological roles for selection theory". It is probably fair to say that most of the pieces are out there, but the topic is far from penetrating the scientific mainstream. It seems as though more work in the area remains to be done.

Saturday, 19 August 2017

The inheritance of acquired sexual preferences

I've long been interesed in the idea that acquired sexual characteristics can be inherited - as part of my more general interest in Lamarckian inheritance. Here is how I have previously described the idea:

The author argues that surgical breast enhancements are inherited, and tend to produce offspring with larger breasts. A mechanism is provided: those with breast enhancements tend to attract mates who prefer larger breasts, and some of that preference will have a genetic basis. Genes in men for a preference for larger breasts will tend to be statistically linked to genes whose expression produces bigger breasts when in women, due to their shared evolutionary history. So: we can expect breast enhancement patients to have offspring with larger breasts than would have been produced if no enhancement surgery had taken place. The reasoning here can be applied to most sexually-selected traits.
I notice that the same logic applies to acquired sexual preferences. A similar example can be used to illustrate this idea. Imagine someone acquires a preference for large breasts - perhaps via exposure to pornography. Their offspring are likely to inherit this preference. How? They are likely to mate with individuals with large breasts, who are in turn more likely than average to carry genes coding for a preference for large breasts.

The fact that the idea also applies to acquired preferences expands its scope. I think that this idea has not been investigated very thoroughly. We don't yet have good theories or models about it. That makes it challenging to judge its overall significance. Another thing that needs doing is empirical testing and quantification. So far, the idea is armchair philosophy. However, the effect should be fairly simple for scientists to measure. It ought to be reproducible with fruit flies or mice, for example. Possibly, data sets suitable for testing the idea may already be out there somewhere.

Defending Lamarck

I think most proponents of cultural evolution acceot the idea that it has a Lamarckian component. I have writen about the topic before - e.g. see: On Lamarckism in cultural evolution. I know many critics accept the role of Lamarckian evolution in culture as well, since one of their refrains is that cultural evolution is not Darwinian, it is Lamarckian.

Lamarck's most famous doctrines these days are the inheritance of acquired characteristics, and the principle of use and disuse. Those are the ideas he is most criticised for holding these days. Textbook orthodoxy says that Darwin's ideas were vindicated while those of Lamarck were rejected. Experiments by August Weismann involving chopping the tails off while mice and observing whether this "acquird characteristic" was inherited are often cited inthis context. The so-caled "Weisman barrier" prevents "acquird characteristics" from finding their way into the DNA of the descendants.

The problem with this is that other traits are inherited. If Weismann had chosen to focus on other traits - such as stress, food preferences or parasite load - he would have found that there was an inherited component. Human examples show the inherited of acquired characteristics most clearly. Jews inherit their missing foreskins from their parents. Tattooed individuals have tattooed offspring. Piercings are inherited. Foot binding, tongue plates, extended necks are all passed down the generations. The inheritance is cultural, not genetic, but Lamarck never confined his views to particular inheritance mechanisms. These were, generally speaking, not known in his day.

There are plenty of examples that don't involve culture too. Dogs inherit their fleas from their parents. Gut bacteria and tooth decay are also acquired characteristics that are inherited. Examples can also be found of Lamarck's principle of "use and disuse". Muscles are a famous example of this principle. With use, muscles grow, and with disuse they shrink. The question is: do offspring inherit their parents muscle distribution? The answer is: yes, sometimes, a bit. The changes are not primarily inherited via DNA - though of course DNA can affect how much you use your muscles. Instead, diet and exercise-related factors that influence muscle size are inherited culturally and through a shared environment.

This is all fairly simple and should be uncontroversial. Nontheless, modern critics of Lamarck refuse to accept that his ideas have any merit. What do they have to say for themselves? Science blogger Jerry Coyne provides a recent example in his article "Aeon tries to revive Lamarck, calling for a “paradigm” shift in evolution". Coyne starts off with a reasonable characterization of the inheritance of acquired characteristics, saying:

Lamarck, of course, was the French biologist and polymath who proposed that animals could stably inherit modifications of their body, behavior, and physiology that were imposed by the environment.
However, then Coyne rapidly goes off the rails, with:

The problem with this idea, and why Lamarck hasn’t become any kind of evolutionary hero, is that it doesn’t work. While the environment can play a role in sorting out those genes that their carriers leave more offspring, there’s no good way for environmental information to somehow become directly encoded in the genome. For that would require a kind of reversal of the “central dogma” of biology

This is, of course a mistaken view. When A mother acquires AIDS, and passes that "acquired characteristic" on to her offsping, no violation of the central dogma is involved. Coyne is totally missing two other possibile ways acquired characteristics can be inherited by offspring: non-DNA inheritance and symbiosis. The idea that DNA modifications must be involved is a very blinkered conception of evolutionary change.

With this, I think, Coyne's critique of Lamarck collapses. The modern vindication of Lamarck doesn't really detract from Darwinian orthodoxy very much. Darwin's ideas still remain very important. I would not describe Lamarckian evolution as much of a "paradigm shift". Darwin himself believed in the inheritance of acquired characteristics, and proposed an elaborate (though mistaken) theory about how they could be inherited. Lamarckian inheritance is more like an extra wrinkle to Darwinian evolution.

Sunday, 13 August 2017

Mutations and recombination in cultural evolution

Another claim in the recent Creanzaa, Kolodny and Feldman document (Cultural evolutionary theory: How culture evolves and why it matters) is my topic today. They say:

Unlike in genetics, where mutations are the source of new traits, cultural innovations can occur via multiple processes and at multiple scales
To start with, this is rather obviously not true: classically, mutations and recombination are the source of new traits in evolutionary theory. However, are the authors correct to claim that these processes need augmenting in cultural evolution? The answer, I think is: not if you conceive of them properly in the first place. Let me explain.

To start with, let's look at what the authors claim are the new processes that go beyond mutation in the cultural domain. They give two examples. One is individual trial-and-error learning. They also say that:

New cultural traits can also originate when existing traits are combined in novel ways
This is cultural recombination - the parallel in cultural evolution of recombination in the organic realm. Do the authors really not know that ideas have sex too?

What about trial-and-error learning, though? Surely there is no leaning in genetics. Trial-and-error learning is a composite process. It starts with trials, which are often mutations of previous trials. Then there is the "error" part, which does not involve generating new variation at all, but rather is based on discarding information based on its success. In other words, it is selection, not mutation or recombination. By breaking trial-and-error learning down into its component parts, it is found to be a composite product of mutation, recombination and selection - not some entirely new process demanding fundamental additions to evolutionary theory. Skinner realised this, by formulating his learning theory while using evolutionary terminology (such as "extinction"). Many others have followed in his footsteps, conceiving of learning in evolutionary terms.

Isn't this a matter of terminology? With these author's definition of 'mutation' they are right, but with my definition of 'mutation', I am right? Yes, but terminology isn't a case of words meaning whatever you want them to mean. Scientific terminology should carve nature at the joints. Definitions of 'mutation' and 'recombination' that apply equally to both organic and cultural evolution are useful, I submit. Less general ones are not so useful.

To summarize, it is possible to conceive of mutation and recombination in a way that make them encompass all sources of variation. Mutations are sources of variation based on one piece of inherited information. Recombination is a source of variation based on two-or-more pieces of inherited information. In theory, it might appear that there's one other possible process: creation - variation based in inherited inforation which comes out of nowhere. One might give the origin of life as an example of genes arising from non-genes. However, we don't really need this proposed 'creation' process. Information never really comes out of nowhere. There's a law of conservation of information - parallel to the laws of conservation of energy and conservation of charge. We can see this in the microsopic reversibility of physics - information is neither created nor destroyed.

I claim then, that mutation and recombination have it covered. The additions to evolutionary theory proposed by these authors are not necessary. They are unnecessaary complications, which evolutionary biologists should soundly reject as not contributing anything to the basic theory.

Saturday, 12 August 2017

Diagram showing where cultural evolution in academia goes wrong

Creanzaa, Kolodny and Feldman have a recent document out titled: Cultural evolutionary theory: How culture evolves and why it matters. It has a nice diagram which is useful in illustrating where academia goes wrong in its study of cultural evolution. Here is the diagram:

The caption reads: "Cultural transmission is more complex than genetic transmission and may occur on short timescales, even within a single generation."

This diagram is profoundly misleading. It is based on a view of cultural evolution that doesn't include symbiology. A genes vs culture diagram that includes cultural symbionts on one side, but not genetic symbionts on the other is not showing the whole picture. Humans share DNA between individuals - in the form of bacteria, viruses, yeasts, fruits and vegetables - very much as they share culture between individuals.

Framing the diagram as "Human genes" vs "Human culture" is not comparing like with like. Bacterial and viral genes are not part of the human genome (unless you count the 10% of the human genome that is descended from viral genomes) - but human culture isn't part of it either. On the left, symbionts are excluded, while on the right they are included. It is an unfair comparison which leads to the confusion propagated by the caption. In fact parasite evolution can happen within a single host generation in both the cultural and organic realms. Contrary to the spirit of the diagram you can get genes from peers in both cultural and organic evolution. They are parasite genes, or symbiont genes in both cases. Cultural evolution does not differ from organic evolution in this respect. The idea that in culture you can get genes from many sources, while in organic evolution you only get them from your parents is a popular misconception about the topic.

The whole document has a whole section on "Culture and Microbes". However there is no mention of the idea that culture behaves similarly to microbes and other symbionts. The man-machne symbiosis, for example is not mentioned. Yet symbiosis is the very basis of the whole field according to memetics, one of the very few symbiosis-aware treatments of cultural evolution out there.

The neglect of symbiology in academic cultural evolution mirrors its neglect in the study of organic evolution - until the 1960s. However, cultural evolution's scientific lag means that cultural evolution is far behind, and few academics have even a basic understanding the relevance of symbiosis to the evolution of culture. Maybe these folk never read Cloak (1975) and Dawkins (1976).

I think the history of this misconception of the whole field in academia is fascinating. Why has it lasted for so long and why has it not yet been corrected? I don't have all the answers but I think the origin is fairly clear. Anthropologists wanted a complex theory of cultural evolution, to signal their skills to other academics and prospective students. They may also have wanted to distance themselves from previous attempts to marry evolution and culture. Any mention of biology turns most anthropologists off. Artificially weakening the influence of biology in the theory may have made the theory more palatable to other anthropologists. Still, science is a self-correcting enterprise. Eventually, the truth will out.