 There are only twenty six letters in the English alphabet. There are also many more types of copied entity than are dreamed of in dual inheritance theory. It follows that those who would continue with terminology along the lines of genes and memes should choose their letters carefully.
I've previously - rather half-heartedly - proposed that we use lemes for learned entities. 'Lemes' cover individual and social learning (by contrast to memes which are normally defined in such a way that they are confined to social learning.
Another proposal which covers similar ground is 'nemes'. The 'n' is short for 'nervous' or 'neural'.
Paul Gilchrist proposed the term 'neme' in 2014 here and here.
A good thing about 'neme' is that it is potentially more inclusive. Not all copying within brains is learning. For example, some is forecasting based on existing models. 'Neme' could plausibly be used as an umbrella term that covers all within-brain copying.
A bad thing about 'neme' is that it is rather closely associated with wetware. Computers also have individual and social learning. However, it seems like quite a stretch to apply the terms 'nervous' or 'neural' to computers.
Another problem is that Paul Gilchrist and I don't seem to agree on what the term should mean. I would want to expand the term to cover copying inside computers and copying of high level structures inside brains - such as ideas. By contrast, Paul says:
I use the term neme to apply to the nerve impulse that is the fundamental element in the operation of the nervous system.
I see where Paul is coming from - but then we need more names to cover all that other stuff that goes on inside brains. We do have 26 letters - but we should use them sparingly and make sure that we don't squander them.
Overall, I quite like the 'neme' term. The surrounding definitional debates show that it needs some more work, though.
 Some say evolution is a theory, others say it is a fact. I tend to regard evolution as a framework.
It is OK to regard evolution as a theory - but as theories go, evolution has a lot of holes in it. On its
own, evolutionary theory doesn't make all that many predictions. It is largely reliant on other
theories to help it to make useful predictions.
It's possible to make a map of the holes in evolutionary theory - to see where other theories
can be attached in order to provide support. That's the main function of this post.
The biggest hole in vanilla evolutionary theory is that it generally lacks a predictive theory
explaining which creatures are fitter. For birds, the additional theories of aerodynamics are
required; for bats, a theory of echolocation is needed - and so on. These other theories are
more closely associated with developmental biology than they are with evolutionary theory.
In general, additional theories that map from genomes to expected fitnesses are required.
Evolutionary theory includes or interfaces to genetics. Genetics also has some holes -
or at least permits other modular theories to be attached to it. Genetics needs
theories of mutation, merging and error correction. There are various types of
mutation: point mutations, frameshift mutations, insertions, deletions and so on;
mutational theories describe what can happen, when it can happen and how likely it
is to happen. "Merging" theories cover recombination, symbiosis and rarer cases where
genomes fuse or assimilate each other. A full theory theory must deal with mate
selection - and the choice of symbiotic partners - since these factors determine
which genomes merge. Error correction theories affect both mutation
and merging. Genome modifications are post-processed by error detection and correction
processes. These bias the results of these processes. Some modifications are permitted,
others are rejected, and others are modified further. Error correction results in
an adaptive bias to mutations - since the most deleterious mutations are selected
against the most strongly by these mechanisms.
Some of the more vocal proponents of the basic idea in this post - that evolutionary theory
contains holes - are Geoffrey Hodgson and Thorbjørn Knudsen. For example, in the paper:
Why
we need a generalized Darwinism, and why generalized Darwinism is not enough,
these authors explain how evolutionary theories do not stand alone and depend on
other bodies of knowledge. I agree with their perspective on this issue.
 In 2003, Geoffrey Hodgson proposed that we use the terms "habit" and "routine" as replacements for the term "meme". As with most other meme synonyms, this suggestion doesn't seem to have been very popular. Retrospectively, it appears to me that this proposal has critical technical limitations that put it out of the running.
Hodgson says that "habits" represent individual transmission while "routines" represent group-level transmission. The dictionary seems to think that individuals can have routines as well, muddying this proposed distinction. Hodgson defends the idea that these entities can act as units of cultural transmission. What he fails to defend is the idea that all cultural transmission is mediated by habits or routines. This claim seems straightforwardly incorrect. For example, the Bible is a bunch of memes, but it isn't a bunch of habits. Habits are associated with individuals, but no individual counts the bible as being among their habits. Nor is the bible a bunch of routines.
This makes Hodgson's proposal incomplete basis of a theory of cultural evolution. If adopting his terminology, we would need one theory for the evolution of habits and routines, and another theory for the evolution of other aspects of culture. Or we would need to redefine these terms and give them counter-intuitive technical meanings. Can we patch up Hodgson's proposal by finding another term (apart from 'habits' and 'routines') to represent other inherited aspects of culture? Maybe - but it looks like a dustbin category to me.
The other issue with Hodgson's proposal is that "habits" and "routines" are not necessarily socially transmitted. We already have terms for mental content that isn't necessarily socially transmitted: 'ideas' and 'concepts'. Part of the reason that term 'meme' found its niche is that it expressed a different idea from the terms 'idea' and 'concept'. If 'meme' had been another synonym for 'idea' and 'concept', it would have failed.
I think Hodgson's proposal is now dead. This post is a post-mortem that attempts to explain where it went wrong.
A cryptic title - but this post is about applying models of universal computation to universal Darwinism.
Most versions of universal Darwinism agree that evolutionary theory applies to brains and thinking. This idea was pioneered by
B. F. Skinner and
D. T. Campbell and promoted by
W. H. Calvin,
G. Cziko and
G. Eldeman among others. Evolutionary theory explains all goodness of fit and all knowledge gain.
If evolution explains the operation of brains, it ought also to explain the operation of computers - since both are general purpose input-transformation-output learning systems. We have some nice, simple models of computation. Can universal computation illuminate Universal Darwinism? In this post we will find out.
We will use the NAND gate + interconnect model of parallel computation and see how it relates to evolutionary models. Copying is a primitive operation in Darwinism: in NAND land it corresponds to signal branching. Selection is another primitive operation in Darwinism: in NAND land, it corresponds to signal termination. That just leaves the NAND gate itself. The NAND gate takes two inputs and produces one output. There are two ways of looking at the NAND operation from a Darwinian perspective. One is as a conditional selection operation. A NAND gate obliterates or inverts one of its inputs depending on the value of the other one. The other way is as a merging or joining operation between two signals. That completes the relationship between these two models.
What did we learn from this exercise? Merging or joining operations turned out to be fundamental. Mutation was not fundamental. It turns out that you can model mutations using copying, selection and merging - if necessary.
Intuitively, the products of evolution include brains - so it is not surprising that some models of evolution are capable of computing partial recursive functions.
However, a universal model has some negative aspects. There's a sense in which universal models are capable of producing any output - and notoriously, models which predict everything are not very useful. We can take some consolation in the idea that there can be all kinds of differences between different universal systems - they differ in speed, degree of parallelism, memory to compute ratio, relative component costs, brittleness, support for synchronous operation - and so on.
One thing I learned from building this model is that my usual reply to critics who allege some Darwinian models lack predictive power is not completely satisfactory. I usually say that constraining the scope of the mutation operator is enough to limit the resulting predictions. However, if there's a recombination operator, that can also lead to universality - and produce a model that is compatible with a lot of observations. It looks as though mutation and recombination both need limiting.
This post presents a model on the level of the bit. Another way of building evolutionary models of computational processes is to rise above the level of the bit. Conventionally, most mutation and recombination takes place between genes - rather than bits - and genes are conventionally quite a bit bigger than bits. This path produces a range of interesting models which have already been well explored in some detail by genetic algorithm and genetic programming enthusiasts.
 I think that most models of the demographic transition have human fertility continuing to fall globally - for some time to come. Recently I read a prediction - posted by Jason Collins - that fertility would rise. Here is Jason's post linking to his article: Fertility is going to go up.
I am pretty sceptical. At one point Jason confesses that he might be wrong, writing:
I’m the first to admit we could be wrong in the prediction of a fertility increase. What other shocks are still to come? Will the continually changing environment drown out the underlying evolutionary dynamics? Our instinct is that most of the shocks that can affect fertility have played out in the developed world – increased incomes, effective contraception, female choice and so on. But what further shocks could reduce fertility?
Here's my attempt at a list of the big fertility-reducing factors that currently still lie largely in the future:
- More engaging games;
- More engaging pornography;
- Sex with robots;
- Economic competition with machines;
- Chemically-induced orgasms;
Basically, memes have run rings around genes, reducing human fertility. IMO, this process shows no sign of stopping - or even slowing down.
The argument that parasites rarely kill their hosts doesn't help much here - parasites can kill their hosts if they have multiple host types and aren't dependent on one particular host type. That will be the situation with intelligent machines - memes won't be dependent on human hosts any more - so they won't be forced to keep them around.
Overall, based on our current understanding of cultural evolution, it seems quite reasonable to model future human fertility as falling to zero. Fertility is going to go down. Jason's argument for the opposite conclusion seems to be based on DNA evolution. However, this is slow - by comparison with cultural evolution. You have to model cultural evolution to have much hope of predicting future changes.
 Part of the muddle still surrounding cultural evolution involves the issue of how to make the genotype/phenotype distinction in the cultural realm.
In the organic realm, this distinction is foundational in genetics, dividing an organism's heritable information from its products. The distinction defines the domain of genetics.
Various authors have written on this issue - including Peter Turchin, Susan Blackmore, Bill Benzoin, Alberto Acerbi and Alex Mesoudi.
Here's what it says about the genotype–phenotype split on Wikipedia:
The genotype–phenotype distinction is drawn in genetics. "Genotype" is an organism's full hereditary information. "Phenotype" is an organism's actual observed properties, such as morphology, development, or behavior. This distinction is fundamental in the study of inheritance of traits and their evolution.
Wikipedia has it right. To me, this seems simple, clear and applies equally well to both the organic and cultural realms. I've been applying this distinction consistently from the beginning - and it works well in cultural evolution.
I can't help but see the position of others on this topic as being muddled and confused. This post raises the question of why this confusion exists. For me, this seems like cultural evolution 101. The genotype/phenotype distinction is a basic issue that people ought to be able to grasp fairly easily. Yet it doesn't seem to be the case. Why are so many researchers in such a muddle?
Perhaps one of the better resources on this issue is Peter Turchin's posts on the topic. They are a bit of a brain dump and illustrate his thought processes at work. However, it does seem to me that most of these authors don't seem to have considered the issue very carefully. For one thing, many of them don't seem to understand why the distinction is important, and apparently think they can get along OK without it. I think this probably indicates a weak background in evolutionary thinking.
Perhaps one problem is that academics often try to get by without the concepts of cultural organisms or cultural creatures. They try to shoehorn everything into the host's genotype and phenotype. If that's the cause of the problem, this issue is just another nail in the coffin of the extended genotype and the host-centric approach to cultural evolution.
 I argue in my observation of the observable video that observation selection effects are part of evolutionary theory - and are simply the result of applying selection theory to observers. I notice that not everyone seems to be on the same page as me about this. Both John Campbell and Matt Ridley have recently ridiculed anthropic reasoning. In chapter one of The Evolution Of Everything, Matt Ridley compares the anthropic principle to a 'skyhook' and writes:
anybody outside a small clique of astronomers who had spent too long with their telescopes, the idea of the anthropic principle was either banal or balmy, depending on how seriously you take it. It so obviously confuses cause and effect. Life is adapted to the laws of physics, not visa versa.
John Campbell doesn't like observation selection effects either. Here are my comments on his treatment in Darwin does Physics, from my book review of it:
I agreed with John in most places - but there were a few areas of disagreement. One was the treatment of observation selection effects. I regard these as a key area where the usefulness of selection is already appreciated by physicists. "Observation of the observable" neatly generalizes Spencer's "Survival of the fittest". However, John's perspective is very different. He says that the anthropic principle is anthropocentric and criticises observation selection as being tautological and unscientific. It is true that the term "anthropic" has an anthropocentric name. However, I think that it is best to ignore the use of this stupid terminology and focus on observation selection effects. These are not tautological or unscientific. Indeed, they help support the case for the applicability of Darwinism to physics - since these selection effects are already well recognised by physicists as a source of 'goodness of fit' between physical law and humans. John has no coverage of this in his book - it seems like a fairly major omission to me.
Both Ridley and Campbell quote from the puddle parable of Douglas Adams to support their ideas. Douglas wrote:
Imagine a puddle waking up one morning and thinking, 'This is an interesting world I find myself in — an interesting hole I find myself in — fits me rather neatly, doesn't it? In fact it fits me staggeringly well, must have been made to have me in it!
Douglas is arguing that reasoning from apparent design to the existence of a creator is not sensible. However this isn't an argument against the utility of observation selection effects! Campbell and Ridley both need to rethink their positions.
 One of the dodgy memes perpetuated by some cultural evolution enthusiasts is that there is more to inheritance than genes. Boyd and Richerson say that human evolution progressed "not by genes alone". Lee Alan Dugatkin calls imitation "evolution beyond the gene". Jablonka and Lamb say that "Genetic" is only one of four dimensions of variation, the others being the "Epigenetic", "Behavioral" and "Symbolic" dimensions. Steven Rose says there is "life beyond the gene". David Sloane Wilson says:
core evolutionary theory needs to expand beyond genetics to include other inheritance systems, such as environmentally induced changes in gene expression (epigenetics), mechanisms of social learning found in many species, and the human capacity for symbolic thought that results in an almost unlimited variety of cognitive constructions, each motivating a suite of behaviors subject to selection (Jablonka & Lamb 2006; Penn et al. 2008).
Expand evolutionary theory without expanding genetics? It makes no sense to me: science needs an expanded genetics too. Expanding the domain of evolutionary theory without expanding the domain of genetics would be a very lop-sided approach.
Memetics pioneered the expansion of genetics to cultural evolution way back in the the 1980s. Efforts to establish a new science of 'replicators' to compete with genetics have gone nowhere - and don't make much sense. The latest attempt to establish a science of non-genetic inheritance - epigenetics - is an absolute joke. What a "dustbin" category that is. It is good for one thing - being an example of how not to do science. Waddington's excellent notion of epigenetics is in the process of being hijacked by foolish and ignorant scientific punks. What we need is a generalized gene and a generalized genetics. When generalizing evolution, scientists should not neglect to generalize genes and genetics! These concepts are absolutely needed for any sensible grounding of evolutionary theory on information theory. Genes should be the units of heredity and genetics should be the study of heredity.
Some say that genetic algorithms are not really "genetic". It is nonsense: genetic algorithms really are genetic. Take genes and genetics seriously, dammit. Don't confine them to the special theory of evolution - that's not where they belong.
Molecular biologists may have appropriated the term "gene" - but it isn't theirs to define. As Steven Pinker puts it:
Molecular biologists have appropriated the term "gene" to refer to stretches of DNA that code for a protein. Unfortunately, this sense differs from the one used in population genetics, behavioral genetics, and evolutionary theory, namely any information carrier that is transmissible across generations and has sustained effects on the phenotype.
If we don't expand genetics now, it will only need doing later. Cultural evolution's scientific lag in academia is large, but forward thinking individuals should still be able to see that the need to generalize genetics is now clear and obvious.
IMO, Boyd, Richerson, Dugatkin, Jablonka and Wilson are not doing scientific progress any favours by dragging their feet on this issue. Get with the program, folks. The 'beyond the gene' meme might look progressive to you - but it looks backwards to me. The best way forwards is to generalize genes and genetics when generalizing evolution. This is scientific evolution, rather than scientific revolution - and evolution is usually less painful and more likely.
Don't tell me that genes and genetics have more inertia and are harder to move. That much is now obvious. The point is that they need to move, and will have to move eventually. So, who is helping? ...and who is not? Which side of this issue are you on?
This one is from November 11th (two days ago). Ridley speaks for the first half, with Ronald Bailey taking over after half an hour.
Matt mentions the "Special" and "General" theories of evolution a few minutes in.
Matt irritates me by repeats]ing an old and ignorant objection to memes by anthropologists. He says (21 minutes in):
You don't have to have particulate information which is what people used to think about cultural evolution that the problem was that you don't have something equivalent to a gene which is a sort of specific, hard object.
The main problem here is the idea that genes are "specific, hard objects". In informational genetics,
genes are informational. They can exist in databases just as well as cells.
My article: Genes are not sections of nucleic-acid goes into more detail.
I feel that the idea that our understanding of cultural evolution has progressed from the idea that culture consisted of "specific, hard objects" to a less "particulate" understanding is a fantasy view of the history of the field which misunderstands and then denigrates the meme enthusiasts.
Memetics is often reductionist - in that it splits cultural inheritance into "bits" of culture. This reductionism is highly productive: a foundational technique in science is splitting complex things into smaller pieces in order to analyse and understand them.
The evolution I would like to see is in people's understanding of genes. Either genes are units of inheritance, or else we need a whole new science of heredity to replace genetics. Keeping genes as units of inheritance is the more conservative and sensible path, I claim. The idea that genes and genetics are confined to the "special theory of evolution" is an awful one. Genetics could be - and should be - quite general. My next post will be all about that topic.
When I got involved with memetics I think that people had previously
used the term "memetics" to refer to the study of memes. It was
more-or-less a synonym for "cultural evolution".
One of the concrete proposals I made was that the split between genetics
and evolutionary theory be extended to cultural evolution - with
memetics mostly studying how memes mutate and recombine. Since
most of the differences between organic and cultural realms
lie in the area, there needs to be a bit of a split between memetics and
genetics - whereas most of the other the areas of evolutionary theory
apply to both domains.
I started promoting the idea that the split between evolution
and genetics should be extended into the cultural domain seriously
in 2013 - with articles like these:
The idea was partly intended as a marketing move for memetics.
Academia has largely opted for the term 'cultural evolution'. However,
according to this terminological scheme, most of the actual
differences between the dynamics of the organic and cultural
realms lie in the genetics/memetics divide. Memetic mutations
and recombination take place inside brains - rather than inside cells -
and so there are expanded opportunities for more complex dynamics.
Most of the alleged difference between the dynamics of the organic
and cultural realms are either differences between memetics and
genetics, fairly direct consequences of those differences, or
are not really differences at all - once symbiology has been
properly taken into account. Differences that turn out to be
insignificant include Lamarckian inheritance, degree of reticulation,
directed mutations and intelligent selection.
This narrative also helps to make sense of the delayed adoption of memetics. In the organic
realm evolutionary theory was discovered in the 1800s, while genetics
didn't really have its own departments until the 1930s. Assuming that
the
study of cultural evolution lags behind the study of evolution
in the organic realm would predict that memetics will similarly
lag behind the study of cultural evolution - an observation that
seems to accord well with the facts.
Another thing I use this narrow version of memetics for is to
disarm critics. For example, critics sometimes assert that
memetics doesn't explain do a good job of predicting why
some memes are fitter than other ones. I sometimes reply to
this that this isn't the job of memetics - any more than it
is the job of genetics to explain why some genes are fitter
than other ones. Just as genetics studies mutation and
recombination, so memetics should study cultural mutation
and recombination. Explaining why one bird's wing works
better than another one would be a job for an aeronautical
expert - not a geneticist. Similarly a memeticist can
reasonably respond that the reason why one bunch of
aeroplane memes has won over another bunch of aeroplane
memes isn't really part of their field of study.
It should be added that some expanded versions of memetics
do get into this topic. Applied memetics, population
memetics and memetic engineering spring to mind as
memetics-related areas that actually do make predictions
concerning meme fitnesses.
Currently I think of the narrow version of memetics
described in this post as being one of my innovations
in the field. However expanding the meme/gene split
into a memetics/genetics split seems like a rather
obvious idea - and I might not be the first one to
take it seriously. In which case my role will have
been to popularize the idea.
I don't have much feedback about the merits of this 'narrow'
version of memetics. Is it the future of memetics? Or is it
a cut-down version of the original vision that just irritates
other practitioners?
References
 This article is a brief commentary on Sylvain Magne's recent article " Objections to Daniel Dennett's informational meme".
Like Dennett, I endorse informational memes. However, Dennett has his own conception of 'information' which differs from that of Shannon. Dennett claims - rather curiously - that there's often no identifiable Shannon channel associated with much hereditary transmission. By contrast, I simply defer to Shannon/Weaver information theory when it comes to the definition of "information". If I want an observer-neutral concept, I simply specify a reference observer.
Sylvain's article rejects the memes-as-information position and proposes Sylvain replace it with memes-as-codes.
I didn't get on with Sylvain's article at all. I have a broader conception of "information" that the one in the paper, and a narrower conception of what constitutes a "code". The term "code" doesn't really have a strict technical definition - but for me a "code" is a collection of symbols - where a "symbol" is something that stands for something else. With the term "code" there's also a pretty strong implication of a symbol-to-symbol mapping. I don't like Sylvain's idea that 'code' is a synonym for 'pattern'. That conflicts too much with common usage. My objection to memes-as-codes is the same as the one Sylvain attributes to a Dennett email in his paper: not all culturally-transmitted phenomena consists of symbols. For example, a wheel is not composed of symbols.
As for my broader conception of "information" - in my defense, my use of the term is the standard one used in information theory used by scientists and engineers everywhere. This is absolutely applicable to heredity - no matter what Daniel Dennett and many other philosophers of biology seem to think.
I point out that academic students of cultural evolution also frequently define their cultural variants in terms of "information". Boyd and Richerson have been doing this pretty consistently since 1985, for example. In fact they even define 'culture' in terms of information. Here's what they said in 1985:
Culture is information capable of affecting individual's phenotypes which they acquire from other conspecifics by teaching or imitation
These days, we would not confine culture to conspecifics or cultural inheritance to teaching and imitation - but that's a rather different topic.
The information-theoretic perspective on heredity has deep roots in evolutionary biology. George Williams pioneered the idea in 1966. He wrote:
In evolutionary theory, a gene could be defined as any hereditary information for which there is a favorable or unfavorable selection bias equal to several or many times the rate of endogenous change
- Williams 1966, page 25 - and later wrote:
A gene is not a DNA molecule; it is the transcribable information coded by the molecule
- Williams 1992, page 11.
An informational memetics thus has strong roots in evolutionary biology and firm foundations in information theory - I claim.
Sylvain mostly critiques Daniel Dennett's conception of "information" - which he says is not clearly specified. That may be so - but I don't think that can be said of my conception of "information". I'm just using the bog-standard scientific and engineering meaning of the term - from Shannon and Weaver.
IMO, if there's a problem with memes-as-information, it is that it is very broad. Saying memes are "cultural" is a lot narrower and confines expectations much more. The idea of informational memes is a rather trivial concept - one which I mostly use to contrast with the position of Aunger - that memes are brain structures. That conception is too narrow to do the work required of it in a theory of cultural evolution.
Appendix
Here is Dennett distinguishing his concept of information from Shannon information (44 minutes in) - as I alluded to earlier in this post.
I'm not talking about bits when I'm talking
about information, I'm talking about information in a more fundamental sense. Shannon information measured in bits is a recent and very important
refinement a one concert with information but it's not the concept I'm
talking about. I'm talking about the concept with
information where when one chimpanzee learns how to
tighten up crack nuts by watching his mother crack
nuts there's information passed from mother to
offspring and that is not in bits that is that is an
informational transfer but has not accomplished in in any
Shannon channel that is worth talking about.
This is a pretty embarrassing quotation from Dennett, IMO. Information which is not measured in bits - pah!
References
|
|
Posts
