Re: Is evolution accelerating?

r norman wrote:
> On Thu, 26 Jan 2006 14:21:16 +1000, John Wilkins <john@xxxxxxxxxxxxx>
> wrote:
>
>> r norman wrote:
>>> On Thu, 26 Jan 2006 11:04:08 +1000, John Wilkins <john@xxxxxxxxxxxxx>
>>> wrote:
....

>>>> The "parent" and "child" relationship is defined by transmission. The parents
>>>> of my allele for blue eyes are of course the alleles in my organismic-level
>>>> parents. So I have organism parents, mum and dad, and gene parents, allele
>>>> blueyes in dad, and allele blueyes in mum. But suppose that the organism
>>>> lineage and the genetic lineage didn't match. Then I would have several
>>>> parent-child lineages. Culture is like that. The "child" of the organisms and
>>>> the "child" of the cultural transmission are *different entities*. I am a
>>>> child of David Hull in the philosophy of science, for example, although he is
>>>> (quite openly) gay and has no biological progeny.
>>> The "parent-child" relationship is necessarily a simply connected
>>> network mathematically. Yes the tree of descent can branch and then
>>> rejoin but there can be no flow of information (genetic material)
>>> traveling in a circle back to the parent. The flow of cultural
>>> information can easily flow in a circle from me to you to many other
>>> people and then back to me. Once you have a circular relationship of
>>> cause and effect, you have the problems of recursion and
>>> self-reference. All sorts of new phenomena arise from exactly this
>>> circularity (yes, I definitely do mean to say emergent phenomena
>>> here). This circularity, this self reference, is the basis of logical
>>> paradoxes. The circularity of information flow is an entirely new
>>> process that can occur in cultural evolution but not in biological.
>> Each case is a new vertex in the graph. The only reason you think it is
>> "circular" is because you identify nodes based on biological location (i.e.,
>> in your own head). But the flow of information is linear.
>>
>> In short, you are *still* bookkeeping culture with respect to organisms.
>
> Yes I "was". See below.
>
> <snip the convoluted context that prompted the next remark>
>
>> Just put your head on this table next to the complete edition of _The
>> Philosophy of the Inductive Sciences_....
>
> Is the game to see which is denser and more impenetrable?
>
> ----------------------------------------------------------------------------------------
>
> Here begins my real posting. But first a couple of meta-notes.
>
> First, it might seem that you have the advantage since you have
> actually pondered the notion of cultural evolution for some time,
> explored it in detail, and even made yourself familiar with what
> others have written professionally and in the primary literature on
> the subject, not to mention having made your own contributions.
> However, the advantage is mine, not only because of my keen intellect,
> but because I am not at all burdened by any of the above baggage!

Sounds like my usual mode of not being hindered by fact sor knowledge. OK, if
it's good enough for me, how can I object to others doing this?
>
> Second, I did in fact assume that cultural ideas were tied to
> individual organisms, people to be specific, just as genes are. I
> realized that fact sometime early this morning while lying in the
> darkness pondering just why I wasn't asleep. But separating out the
> memes, or whatever you want to call them, as disembodied entities
> floating "out there" in the aggregate mental space of humanity doesn't
> help: see the real content below.
>
> I now propose two (insurmountable) objections to the notion that
> cultural evolution parallels biological evolution in anything much
> more (note the careful hedge) than metaphor.

OK, this sets the stage. Let's see how you do.
>
> First, the fact is that the subject matter of biological evolution,
> genes (more properly, the genetic composition of a population), has a
> physical reality while the subject matter of cultural evolution does
> not. This, in fact, is exactly the point you raised about my

The notion of what a gene is, particularly with respectt to evolution, is very
far from having a single, univocal and mutually consistent definition or
physical substrate. Here is a paper that describes the state of play:

http://philsci-archive.pitt.edu/archive/00002494/

see also

http://www.nature.com/embor/journal/v6/n9/full/7400521.html

It is easy to think things are more simple than they are WRT genes. Are you
talking about Mendelian factors, evolutionary genes, ORFs, expressed primary
sequences edited at the transcription stage?

> mistakenly "bookkeeping culture with respect to organisms." Genes
> (admittedly a difficult word used in different senses by different
> people, but you know what I mean) originally were associated somewhat
> vaguely with "traits" that could be inherited. But molecular biology

So, Mendelian factors...

> has in fact demonstrated that they are also associated with specific
> sequences of DNA that are connected with those traits (albeit often in
> a very circuitous and sometimes non-deterministic way). That means it

Oh, you mean molecular gene concepts? Unfortunately these do not map onto the
first kind easily or uniquely.

> really is possible to actually measure the genetic composition of a
> population experimentally. Conceptually all you have to do is to
> sequence the genome of every member of the population. Practically,

That all? :-) But in fact that is never done, and from the theoretical
perspective it remains undoable. So instead you sample populations according
to some prior assay or target sequence. And you provide accounts of how the
frequencies inferred by these samples occurred.

> population genetics has given us tools to make excellent estimates of
> gene frequency. For example, every intro biology students who comes
> into contact (intellectually, not physically) with me has gone through
> the exercise of counting the fraction of the class population with
> detached earlobes, tongue curling ability, PTH tasting, etc. and then
> computed the frequency of the recessive allele as the square root of
> the frequency of the recessive phenotype. It is a silly exercise to
> illustrate Hardy-Weinberg algebra (although no one ever mentions that
> the class is not a population in H-W equilibrium) but it does
> emphasize the point that, given appropriate conditions (assumption of
> a near equilibrium and knowledge of pattern of heritablility) it is
> indeed possible to measure something not directly observed (gene
> frequencies) from something that is observed (phenotype frequencies).;

Sure. But can you demonstrate this is not possible with memes?
>
> Now compare "cultural ideas" or "memes" or "ideas" or whatever they
> are. They exist, as far as we know, only in our mind. We have
> absolutely no idea what that means in terms of our physical brain.

Well, that's not entirely true on two counts. First, we are getting a better
resolution of the ways in which brains instantiate thoughts and where and how.
Second, I do not think it is true that memes exist only in heads. Any meme
worth its salt exists as an alternative in the *cultural* context. A colour
name, for instance, is an "allele" in colour "space" that exists in the
lnaguage community. "Thinking about 'red'" is something you do as an English
speaker, not as a physical system in thus and so activation pathways. But we
will learn about those activation pathways the way we learned about how
hereditable traits had a molecular story underlying them.

> And they are evanescent, flickering in and out like vapors. We can
> hold completely contradictory ideas simultaneously, variously acting
> upon first the one then the other. We can change our minds in an
> instant. We can pretend to believe one thing when asked but really do
> something else.

You set up the strawman in order to knock it down here. I do agree it is hard
to find something solid underlying cultural items (I published a paper on it
recently objecting to an attempt to identify memes with neural states), but it
doesn't follow thereby that memes are evanescent. If I say to you "Cat!" and
point, then no matter what the underlying substrate might be in you or me or
the three year old down the road past whom it is running, there is something
rather *not* transient about that meme.

You are confusing here the type with the tokens. In *your* mind or mode of
cultural expression a thought may be fleeting. But the type is persistent. The
whole enterprise I have been engaged in on talk.origins for the past 15 years
has been to try to engage a persistent misunderstanding about evolution. This
is not evanescent - it has survived 200 years at least. You might briefly
entertain a notion about evolution that derives from this meme, and lose it
because you are better informed than Gish or Morris. But the meme itself persists.

> Yes, organisms can alter their phenotype and can even
> lie about it (as in mimicry) but the process tends to be rather slow
> and stodgy, not instantaneous as with ideas. Worse, the very act of

Again, you conflate the organism having the ideas with the ideas as a type.
It's that level confusion coming in again. *Memes* don't fleetingly alter
their "phenotype" (I called it once a "phemotype" but it didn't take); that
persists in the culture. And why should the transience of a phenotype be an
argument against it evolving? Even *if* the phenotype of, say, a virsu,
persists no longer than a few moments, it is persistent in the evolution of
that "organism" over its evolutionary time.

> measuring or observing somebody's opinion can in fact change it. This
> is something far in excess of the physical process in which a
> measurement alters the measured based on Heisenberg uncertainty. That

Here you confuse observation with the thing observed (a common problem with
empirically-oriented biologists, I note). The *operational* difficulty of
measurement and tracking is not an argument against the actuality of its
evolution. For example, every time you sacrifice an organism to assay it - to
disecct it or even sample its genotype, you affect the dynamics of that
population. You just expect that the effect will be well within the error bars.

> uncertainty is, in fact, a very precise expression between two very
> definable features. Measuring position does not totally alter the
> experiment, it just renders uncertain the momentum. Measuring ideas
> alters them in ways completely unpredictable. You might then be left
> with examining, not the ideas, themselves, but physical traces that
> they may leave -- studying jet aircraft by looking at their contrails.

I do think that you can track the "epidemiology" of ideas in ways that do not
affect the trajectory of them, particularly at a relatively high level and
after the fact. For example, most social history does this already. We do not
need, I think, to have a new methodology, although that is clearly a good
thing if we can. We need to have an account of how culture evolves that
prevents us inserting our research biases. Treating things as darwinian
entities the way I would like (I'm not altogether happy with the term "meme"
because it relies on what I think is a problematic ontology of evolution)
enables us to ask the right research questions.

> The connection between gene and phenotype is something that used to be
> rather difficult; the trend now is to look at genetic diversity
> directly by sequencing DNA and that is absolute. There is no
> equivalent in cultural evolution. The connection between the specific
> nature of the ideas and the physical traces they leave is far too
> ambiguous except in very special cases. I don't even know that you
> can ascribe a reality to the notion of an "idea frequency" and you
> will have to convince me that there really are good experimental
> methods to measure it.

There was no absolute measure for genes until relatively recently. This
attitude would have stopped genetics dead in its track in 1900.

But we can identify tokens of memes and track their frequency and spread well
enough to get going.
>
> In addition, the connection between gene and organism is even more
> fundamental than the physical reality of the gene. The "measure" that
> selection works under is something called "fitness". We go to great
> lengths to explain that fitness is not a circular notion -- what is
> fit is not merely what gets selected. Fitness is, in fact, closely
> tied to the fact that it influences the organism that possesses that
> genome; its longevity, its fecundity, its parental success. And
> fitness is something that is stable -- once you are endowed with that
> genome, that is what you keep through your reproductive years.
> Further, your offspring will obtain a genome that is clearly derived
> from your own (and your partner's). With ideas, this connection no
> longer applies. What is the measure of ":fitness" apart from its
> success in getting passed on? Can you demonstrate the non-circularity
> of that definition? Can you be confident that the mechanism of
> transmission (its reproduction) or its survival has anything
> whatsoever to do with the "fitness" of the idea as opposed to the
> fickleness of its possessor, a trait that is in no way correlated with
> the idea, itself? Can you even insure that the offspring of an idea
> has much connection with the original idea? If every reproductive
> step involves so much mutation that children only vaguely resemble
> their parents, then we would have a difficult time recognizing
> evolution.

Yes to all these. Fitness of memes is a function of the social environment and
we can give, at least in theory, an account of why a meme is fitter in America
than in Uganda in terms of the social economy (that is, in terms of the social
resources needed to spread the meme). This will rely on the social structure,
the structure of cultural institutions, and so ofrth. I find that a game
theoretic analysis often makes it clear what the resources are, in terms of
the payoffs that inform the decision matrix.
>
> I will agree that there are aspects of cultural evolution that work
> slowly enough -- the structure of language, of religious ritual

Stop! - "slowly" is a relative term. What is slow for mammalian evolution is
different for what is slow for viral evolution. This is immediately biasing
the issue in ways that make it hard for culture to be seen as an evolutionary
process - it's question begging.

But if you regard things in terms of generations (not biological but cultural
generations) and the relative probability that the meme will be transmitted
accurately, then you have the sort of thing that is "slow" enough, "gradual"
enough to have an evolutionary process.

> associated with life events, most of the traditional material of
> anthropologists -- so that there really is a close connection between
> ideas and individuals. The ideas tends to remain fixed for years and
> years and societies, at least pre-industrial societies, had cultural
> transmission mechanisms that were very conservative. In those cases
> you might make a reasonable comparison between cultural and biological
> evolution. Even then, there were problems. How valid were Margaret
> Mead's notions of sexuality among the Samoans? Modern life has made
> communication of ideas a radically different story. Ideas can spread
> instantaneously (nearly) around the world. Are "pet rocks" and "hip
> hop slang" in the same league as the traditional "marriage taboos" or
> "dropped initial aspirated 'h' sound"? Even if cultural evolution
> theory based on biological evolution is successful in certainly
> clearly defined areas, is that enough to build a general theory?
>
> In short, are the fundamental elements of cultural evolution
> "entities" that can be defined and measured scientifically?

I do not think so. Neither do I think there are universal fundamental units in
all cases of biological evolution. Sure, sometimes point mutations are the
unit of evolution - sometimes regulatory genes, sometimes expressed genes.
Sometimes entire populations, kin groups or demes are. Sometimes traits evolve
independently sometimes not.

But we can do with culture what we do with biology and measure things that
occur recognisably in causal chains and identify the frequencies with which
they occur, and hypothesise selective pressures or drift, and test them.
>
> The second point has to with your utter rejection of emergent
> properties. It is this obstinate reductionism that causes you to
> argue that, since ideas are the product of the human brain and since
> the human brain is the product of evolution, then ideas (hence
> cultural evolution) are the product of evolution. There is a shred of

Obstinate reductionism? Au contraire, I often think things happen at a level
that is independent of the lower constituent levels. In fact in this case I am
arguing that ideas et al evolve independently of biology *because* they are
not reducible to genetic or biological models.

> truth in those statements -- they are each true. But they don't lead
> to the conclusion that, therefore, cultural evolution is just an
> extension of biological evolution. As an argumentum ad absurdum (or

I do not make that inference. And the Latin is reductio ab absurdum.

> whatever the Latin is), all of biology is chemistry is physics. So
> therefore I should be able to solve the Schrodinger equation for a DNA
> molecule to determine the fitness of the resulting organism in any
> given environment. It might be a difficult and lengthy computation
> but it is, in fact, doable in concept. No it is not.

No, it is not. This is because not all instances of physics will exemplify all
possible states of the Schroedinger equation - what is be physically possible
is a larger set than what is actual. You are making a strawman and calling it
John Wilkins.
>
> Once the physics and chemistry of a system reaches a certain level of
> complexity, the system adopts a completely new pattern of dynamics.

Completely new? You mean there are things that do not obey the laws of physics?

> In fact, there can be a sharp threshold in systems between those which
> show the new dynamics and those which do not. The new dynamics, the
> new set of solutions to the system equations, is called "life". That
> is, "life" is a physical system whose dynamic trajectory (in a
> suitably defined mathematical phase space) is confined to a particular
> basin of attraction (a technical term) called the "living state".
> Having that basin is, in fact, exactly the new dynamics not possessed
> by an inanimate system. And the biological system (one which
> possesses such a basin) can in fact, have a trajectory that leaves the
> basin. In that case, the system dies. The point is that there are
> completely new concepts: "life" and "death" possessed by such a system
> that are not possessed by a slightly different though apparently
> almost equivalent dynamic system. And, when you get to a
> self-reproducing system with an imperfect information copying system,
> you get "factors" or "genes" or "informational segments" that can
> mutate and influence fitness and produce evolution. In other words, I
> claim that evolution is an emergent property of such a biological
> system of replicating organisms with imperfectly heritable features
> that determine ability to live, function, and reproduce and pass on
> those features to the next generation.

You commit, since we are using classical logical terms, the fallacy of
composition, or division. I do not say that the properties of the parts must
be the same as the properties of the whole. I just say that there is a formal
isomorphism between the ways living systems evolve and the way cultural
systems evolve. If we construct, as we have in ALife, a mechanical system that
did not evolve (say, for purposes of argument, it flashed into existence via
the Infinite Improbability Drive), and it met the preconditions for evolution
listed in my other post, then the "entities" in that system would evolve
darwinianly. Not because the components evolved, but because the entities now
in existence have no other possibility.
>
> I now claim that a system that allows ideas to be generated and to
> become copied from one organism to another albeit in different form
> has crossed another threshold into a new dynamic system called
> "culture". In this new scheme, there are completely new properties
> that are quite reminiscent of those of the underlying biological
> system we call evolution. However, these new entities have their own
> properties and, especially, their own dynamics so that the theory of
> cultural evolution is, indeed, quite a different creature from the
> theory of biological evolution.

They do have their own supervening properties. And those properties are the
general ones that permit, indeed require, them to evolve in darwinian fashion.
>
> In no way do I deny the existence of cultural evolution or that
> interesting studies can be made of its processes. I only claim that
> it bears only passing resemblance to biological evolution.
>
> So there!

Back at you.

--
John S. Wilkins, Postdoctoral Research Fellow, Biohumanities Project
University of Queensland - Blog: evolvethought.blogspot.com
Servum tui ero, ipse vespera

.

Relevant Pages

  • Re: Is evolution accelerating?
    ... So I have organism parents, mum and dad, and gene parents, allele ... >> process that can occur in cultural evolution but not in biological. ... every intro biology students who comes ...
    (talk.origins)
  • Re: Is evolution accelerating?
    ... >>> that such culture is separate and apart from their biological evolution. ... >> cultural evolution must be treated separate from biological. ... > out to be irrelevant to the main dynamics of cultural evolution. ... you can't ask if culture evolves more rapidly than biology ...
    (talk.origins)
  • Re: Evolution and Love
    ... cultural evolution and biological evolution. ... traumatizing lifetime conditions and events - into the brains of very ... reasoning and a metaphor for the Actention Selection System, ...
    (sci.bio.evolution)
  • Cultural evolution and biological evolution
    ... Wallace and Charles Darwin, and made famous by Darwin's book of 1859, ... The Origin of Species. ... Yet biological and cultural evolution each have ... Already we can see a new way in which cultural evolution is ...
    (sci.bio.evolution)
  • Re: Are you Conscious?
    ... without relying too much on the precise sets of genes and neuronal ... Sociobiology and evolutionary psychology take ... Genetics and molecular biology have provided some significant insights ... alot more leeway for their theories of human evolution. ...
    (sci.bio.evolution)