Felsenstein v. Dembski
John Harshman
issues of the Reports of the NCSE. But volume 27 #3-4 contains a great
article by Joe Felsenstein that neatly demolishes every one of William
Dembski's arguments in easy to understand language. I recommend it to
anyone, but especially Tony.
Ah, I see it's on the web too:
http://ncseweb.org/rncse/27/3-4/has-natural-selection-been-refuted-arguments-william-dembski
wf3h
wrote:
thanks for the update...very informative. i was shocked to read
stephen meyer being quoted saying:
"We know that information — whether, say, in hieroglyphics or radio
signals — always arises from an intelligent source. .... So the
discovery of digital information in DNA provides strong grounds for
inferring that intelligence played a causal role in its origin. (Meyer
2006)"
because this seems to indicate ALL of nature is 'intelligently design'
since all of nature contains SOME type of information...hell even
crystals have information on orientation, periodicity, etc.
Perplexed in Peoria
Yes, nice. Thx. And it led me to reread Wolpert's anti-Dembski piece
http://www.talkreason.org/articles/jello.cfm
Also nice, and after reading Felsenstein I see that it does a bit more
than simply insult Dembski's clarity.
Bob Casanova
in talk.origins, posted by John Harshman
<jharshman....@pacbell.net>:
Great; thanks!
--
Bob C.
"Evidence confirming an observation is
evidence that the observation is wrong."
- McNameless
John Harshman
argument against Dembski's appropriation of NFL, i.e. that the fitness
surfaces of living organisms (and genome, and proteomes) are not a
random sample of all conceivable fitness surfaces, but surfaces in which
near neighbors are typically highly correlated, i.e. those in which
hill-climbing algorithms are expected to do quite well. Felsenstein
makes this point with devastating force.
VoiceOfReason
Ray Martinez
wrote:
"Mutations
In the real world, mutations do not act like this. Yes, they are much
more likely to reduce fitness than to increase it, but many of them
are not lethal. I probably carry one — I have a strong aversion to
lettuce, which to me has a bitter mineral taste. This is probably a
genetic variation in one of my odorant receptor genes. It makes salad
bars problematic, and at sandwich counters I spend a lot of time
scraping the lettuce off. But it has not killed me — yet. The great
body of empirical information about the effects of mutation in many
organisms makes it clear that a great many mutations are not instantly
lethal. They do on average make things worse, but they do not plunge
us instantly back into the primordial organic soup."
I stress:
"They do on average make things worse"
Not a word about positive, good or beneficial mutations, that is, the
ones that an alleged unguided material process somehow knew to
preserve.
Ray
wf3h
did you read his calculation of a very very minor beneficial mutation
can, over 84 generations, come to be present in over 99% of the
population?
it doesn't take much at all.
Ray Martinez
BEGIN QUOTE
"Dembski himself seems unable to draw this self-evident conclusion
from his own argument. He acknowledges that "the development of
antibiotic resistance by pathogens via the Darwinian mechanism is
experimentally verified and rightly of great concern to the medical
field" (Dembski 2002: 38). But by saying that he undercuts his own
argument — if correct, his argument would actually prove that the
adaptive information in the bacterial genome could not be created by
natural selection, except by the pure accident of mutation and genetic
drift, unaided by natural selection."
END QUOTE
But the Explanatory Filter allows for the concepts seen in the
Darwinian mechanism----that's why Dembski said what he said. Dembski
is only denying the main positive claim of natural selection (creative
force). And I wouldn't be surprised if Sean Pitman showed up here in
this context. As far as I can tell, Sean advocates both unintelligent
and Intelligent agencies to be operating in reality. Of course the
History of Science doesn't allow this choice. Its one or the other. We
(= Biblical Creationism) can explain antibiotic resistance within our
supernatural paradigm.
Ray
Dana Tweedy
> On Mar 23, 8:41 am, John Harshman <jharshman.diespam...@pacbell.net>
> wrote:
>> For reasons too complicated to explain, I am only now reading
>> 2-year-old issues of the Reports of the NCSE. But volume 27 #3-4
>> contains a great article by Joe Felsenstein that neatly demolishes
>> every one of William Dembski's arguments in easy to understand
>> language. I recommend it to anyone, but especially Tony.
>>
>> Ah, I see it's on the web too:
>>
>> http://ncseweb.org/rncse/27/3-4/has-natural-selection-been-refuted-ar...
>
> "Mutations
> In the real world, mutations do not act like this. Yes, they are much
> more likely to reduce fitness than to increase it, but many of them
> are not lethal. I probably carry one — I have a strong aversion to
> lettuce, which to me has a bitter mineral taste. This is probably a
> genetic variation in one of my odorant receptor genes. It makes salad
> bars problematic, and at sandwich counters I spend a lot of time
> scraping the lettuce off. But it has not killed me — yet. The great
> body of empirical information about the effects of mutation in many
> organisms makes it clear that a great many mutations are not instantly
> lethal. They do on average make things worse, but they do not plunge
> us instantly back into the primordial organic soup."
>
> I stress:
>
> "They do on average make things worse"
The point was that all mutations are not fatal, or even particularly
harmful. The vast majority of mutations are neutral.
>
> Not a word about positive, good or beneficial mutations, that is, the
> ones that an alleged unguided material process somehow knew to
> preserve.
The unguided processes don't "know" to preserve the beneficial mutations.
They do preserve them because the organisms that have the beneficial
mutations are more likely to leave more offspring.
DJT
Dana Tweedy
> On Mar 23, 8:41 am, John Harshman <jharshman.diespam...@pacbell.net>
> wrote:
>> For reasons too complicated to explain, I am only now reading
>> 2-year-old issues of the Reports of the NCSE. But volume 27 #3-4
>> contains a great article by Joe Felsenstein that neatly demolishes
>> every one of William Dembski's arguments in easy to understand
>> language. I recommend it to anyone, but especially Tony.
>>
>> Ah, I see it's on the web too:
>>
>> http://ncseweb.org/rncse/27/3-4/has-natural-selection-been-refuted-ar...
>
> BEGIN QUOTE
> "Dembski himself seems unable to draw this self-evident conclusion
> from his own argument. He acknowledges that "the development of
> antibiotic resistance by pathogens via the Darwinian mechanism is
> experimentally verified and rightly of great concern to the medical
> field" (Dembski 2002: 38). But by saying that he undercuts his own
> argument — if correct, his argument would actually prove that the
> adaptive information in the bacterial genome could not be created by
> natural selection, except by the pure accident of mutation and genetic
> drift, unaided by natural selection."
> END QUOTE
>
> But the Explanatory Filter allows for the concepts seen in the
> Darwinian mechanism----that's why Dembski said what he said.
Of course, Dembski's "explanitory filter" is useless in the real world.
> Dembski
> is only denying the main positive claim of natural selection (creative
> force).
Of course, natural selection is a creative force when coupled with random
mutations.
> And I wouldn't be surprised if Sean Pitman showed up here in
> this context. As far as I can tell, Sean advocates both unintelligent
> and Intelligent agencies to be operating in reality.
As does anyone who understands the situation. Realizing, of course that
the "intelligent agencies" are human beings, at least as far as can be
determined. There may be other intelligent agencies, but if they exist,
they are undoubtedly naturally occuring beings.
> Of course the
> History of Science doesn't allow this choice.
Of course, Ray is ignorant of the history of science.
> Its one or the other.
Which is the primary evidence for my previous claim. Anyone who is aware
of the evidence knows that both non intelligent forces exist, and that human
intelligence is seen in nature.
> We
> (= Biblical Creationism) can explain antibiotic resistance within our
> supernatural paradigm.
Of course a "supernatural paradigm" can explain anything, which is why it's
not science. There's no way to falsify appeal to the supernatural. It's
useless as an explanation.
DJT
John Harshman
As usual, you don't understand the argument. Felsenstein isn't talking
here about the Explanatory Filter. He's talking about No Free Lunch. If
Dembski's application of NFL to evolution were correct, natural
selection would be impossible because nearly all mutations would be
lethal. The fact that Dembski doesn't realize the implications of his
argument is not in his favor.
> And I wouldn't be surprised if Sean Pitman showed up here in
> this context. As far as I can tell, Sean advocates both unintelligent
> and Intelligent agencies to be operating in reality. Of course the
> History of Science doesn't allow this choice. Its one or the other. We
> (= Biblical Creationism) can explain antibiotic resistance within our
> supernatural paradigm.
Do you have a tapeworm again?
wf3h
you're misreading what dembski is quoted as saying.
he tries to have it both ways. while he is forced to admit the
EXPERIMENTAL evidence of 'micro' evolution...because it's readily
seen, his mathematical analysis PROHIBITS even the simplest cases of
'micro' evolution because it requires the complete independence of the
entire structure of the genome. as the author says, it's as if when
you close your window, your newspaper stops being delivered.
it's a fatal objection to dembski's argument, also made by wolpe, in
the sense that dembski's assumption is that co-evolution does not
exist. it plainly does, dembski admits it does (as does pitman) and
this small admission kills dembski's analysis.
wf3h
wrote:
in addition, dembski excludes co-evolution in his analysis. as
felsenstein points out, this would be analogous to the situation when
you close your kitchen window to go on vacation, your neighbor forgets
to watch your house in spite of a previous arrangement.
dembski is forced to admit microevolution exists (bacterial
resistance) but his view of the NFL theorem would make the development
of bacterial resistance impossible.
Perplexed in Peoria
>>> For reasons too complicated to explain, I am only now reading 2-year-old
>>> issues of the Reports of the NCSE. But volume 27 #3-4 contains a great
>>> article by Joe Felsenstein that neatly demolishes every one of William
>>> Dembski's arguments in easy to understand language. I recommend it to
>>> anyone, but especially Tony.
>>>
>>> Ah, I see it's on the web too:
>>>
>>> http://ncseweb.org/rncse/27/3-4/has-natural-selection-been-refuted-arguments-william-dembski
>>
>>
>> Yes, nice. Thx. And it led me to reread Wolpert's anti-Dembski piece
>> http://www.talkreason.org/articles/jello.cfm
>> Also nice, and after reading Felsenstein I see that it does a bit more
>> than simply insult Dembski's clarity.
>>
> I also think that Wolpert, paradoxically, fails to make the most telling
> argument against Dembski's appropriation of NFL, i.e. that the fitness
> surfaces of living organisms (and genome, and proteomes) are not a
> random sample of all conceivable fitness surfaces, but surfaces in which
> near neighbors are typically highly correlated, i.e. those in which
> hill-climbing algorithms are expected to do quite well. Felsenstein
> makes this point with devastating force.
Yes he does - the bit about architecture training was particularly cute.
The world we live in is not encrypted. Most parts of it interact very little
with other parts. When my family leaves home for a vacation, we have
to make many arrangements at home concerning doors, windows, lights,
toilets, faucets, thermostats, garbage, notifying neighbors, stopping delivery
of newspapers, and so on. If we lived in Dembski's encrypted universe,
this would be impossible. Every time we changed the thermostat setting,
the windows would come unlocked and the faucets would run. Every time
we closed a window, the newspaper delivery would resume, or a neighbor
would forget that we were leaving. (It's worse than that, in fact. The house
would be totally destroyed.) But, as we live in the real universe, we can
cheerfully set family members to carrying out these different tasks without
their worrying about each other's actions. The different parts of the house
scarcely interact.
Of course a house is a designed object, but it is not particularly hard to
have its parts be almost independent. When architects train, they do not
have to spend much of their time ensuring that the doors, when closed,
will not cause the faucets to run.
But Dembski might claim (and Felsenstein might concede) that the
sheer fact that our universe is one in which that kind of modularity
is there "for free" is itself something of an argument for the "fine-tuning"
version of ID. So now I need to understand why it is probably the
case that modularity is natural in most universes in the multiverse. I
think it is, and I think I know why, but I'm not sure I can express it.
Felsenstein also suggests the possibility that modularity can itself
evolve under NS. Evolution of evolvability. That one makes me
nervous, but it is surely possible in a multi-level selection environment.
Trouble is, in this context, it simply re-raises Wolpert's problem
of induction. What is it that guarantees that a mutation that makes
one characteristic more evolvable will tend to make all characteristics
more evolvable?
Architects don't spend a lot of time learning how to let in light without
letting in rain - they put the windows on the walls rather than ceilings
because their predecessors did. They don't spend a lot of time learning
how to let in light without letting out heat. They glaze the windows
because glass is implied by the word "window" in modern societies.
And they don't spend a lot of time worrying about how to let people
in and out without having the whole structure collapse. Nor did I
when I recently built a shelter from brush and twine. But it collapsed
anyways.
John Harshman
I'm confused here. Felsenstein's example has nothing to do with
coevolution. Several people have made much of coevolution invalidating
Dembski's ideas, but as Felsenstein points out, there are much more
basic and general problems; there's really not much reason to bring up
coevolution.
> dembski is forced to admit microevolution exists (bacterial
> resistance) but his view of the NFL theorem would make the development
> of bacterial resistance impossible.
That's right, as Felsenstein says.
John Harshman
what is it, specifically the one that makes sense of this post?
John S. Wilkins
Must be good. Practically the first thing he cites is mine and Wesley's
paper.
--
John S. Wilkins, Philosophy, University of Sydney
scienceblogs.com/evolvingthoughts
But al be that he was a philosophre,
Yet hadde he but litel gold in cofre
wf3h
agreed, but felsenstein discusses it both in terms of his argument and
in terms of wolpe's objections to dembski's argument
>
Ray Martinez
wrote:
Or you don't.
> Felsenstein isn't talking
> here about the Explanatory Filter. He's talking about No Free Lunch.
Agreed.
But Dembski IS by silent presupposition since NFL exists in the
context of previous work (EF). Just like Darwin 1871 exists in the
context of Darwin 1859. I have explained why Dembski made the comment.
The EF allows the main concept seen in the Darwinian mechanism to
explain reality if it meets the criteria.
> If Dembski's application of NFL to evolution were correct, natural
> selection would be impossible because nearly all mutations would be
> lethal. The fact that Dembski doesn't realize the implications of his
> argument is not in his favor.
>
Actually I don't have a problem with Felsenstein's point/observation.
It supports my view that natural selection does everything claimed or
does nothing at all. I am simply anticipating what Dembski might say
since I have seen Sean Pitman make the same point when "cornered."
Ray
> > And I wouldn't be surprised if Sean Pitman showed up here in
> > this context. As far as I can tell, Sean advocates both unintelligent
> > and Intelligent agencies to be operating in reality. Of course the
> > History of Science doesn't allow this choice. Its one or the other. We
> > (= Biblical Creationism) can explain antibiotic resistance within our
> > supernatural paradigm.
>
> Do you have a tapeworm again?- Hide quoted text -
>
> - Show quoted text -
wf3h
wrote:
>
see wolpe. it's central to wolpe's objection to dembski
\
John Harshman
Who is Wolpe and where do I find him?
John Harshman
We agree on this. But the point, which you seem to be unable to notice,
is that Dembski's interpretation of NFL doesn't. Either Dembski is wrong
or natural selection is impossible, though he doesn't realize it.
>> If Dembski's application of NFL to evolution were correct, natural
>> selection would be impossible because nearly all mutations would be
>> lethal. The fact that Dembski doesn't realize the implications of his
>> argument is not in his favor.
>
> Actually I don't have a problem with Felsenstein's point/observation.
> It supports my view that natural selection does everything claimed or
> does nothing at all. I am simply anticipating what Dembski might say
> since I have seen Sean Pitman make the same point when "cornered."
No, it doesn't support your view. And that may indeed be what Dembski
might say, since he doesn't get it either, and tends to avoid responding
to expert criticism.
>>> And I wouldn't be surprised if Sean Pitman showed up here in
>>> this context. As far as I can tell, Sean advocates both unintelligent
>>> and Intelligent agencies to be operating in reality. Of course the
>>> History of Science doesn't allow this choice. Its one or the other. We
>>> (= Biblical Creationism) can explain antibiotic resistance within our
>>> supernatural paradigm.
>> Do you have a tapeworm again?
Or some other internal parasite?
Ernest Major
<d12a0a7a-8141-4547...@w34g2000yqm.googlegroups.com>,
wf3h <wf...@vsswireless.net> writes
exist.
The No Free Lunch theorem boils down to *on average*, *over all
problems*, no search algorithm can perform better than random search.
Contrary to Dembski's claims that doesn't mean that there are no search
algorithm that can perform better than random search is relevant
domains. For example if Dembski was right the Newton-Raphson method
would not work.
An analogy that I offered earlier is data comprehension. It is another,
and possibly related, theorem of information theory that no lossless
compression algorithm can compress all strings (in fact the pigeonhole
principle implies a stronger statement should be possible). But that
doesn't mean that lossless data compression is impracticable in practice
- see the existence of ZIP, GIF, PNG etc files for counterexamples.
--
alias Ernest Major
Ernest Major
<jharshman....@pacbell.net> writes
cited at the top of the thread.
As I understand the matter, the relevance of coevolution is that the No
Free Lunch theorem applies to searching *fixed* spaces - but coevolution
means that the spaces involved change over time.
--
alias Ernest Major
John S. Wilkins
Just out of interest, how is the Designer supposed to have found the
solutions, then? Is design just random search after all?
>
> An analogy that I offered earlier is data comprehension. It is another,
> and possibly related, theorem of information theory that no lossless
> compression algorithm can compress all strings (in fact the pigeonhole
> principle implies a stronger statement should be possible). But that
> doesn't mean that lossless data compression is impracticable in practice
> - see the existence of ZIP, GIF, PNG etc files for counterexamples.
--
TomS
<1ix3ryn.1vizy7at2er4nN%jo...@wilkins.id.au>, John S. Wilkins stated..."
[...snip...]
>
>Just out of interest, how is the Designer supposed to have found the
>solutions, then? Is design just random search after all?
What I'm curious about is the *problems* that the Designer had to
cope with.
Presumably, complex solutions are an indication of problems which
were complex for the solver (namely, the designer).
But, even if the problems were simple, where did they come from?
--
---Tom S.
"As scarce as truth is, the supply has always been in excess of the demand."
attributed to Josh Billings
Perplexed in Peoria
>> The No Free Lunch theorem boils down to *on average*, *over all
>> problems*, no search algorithm can perform better than random search.
>> Contrary to Dembski's claims that doesn't mean that there are no search
>> algorithm that can perform better than random search is relevant
>> domains. For example if Dembski was right the Newton-Raphson method
>> would not work.
>
> Just out of interest, how is the Designer supposed to have found the
> solutions, then? Is design just random search after all?
The Designer can solve the search problem by Brute Force. Only we
finite intellgences need to use computation and algorithms.
Hmmm. It appears that ID can tell us something about the Designer
after all. ...
Burkhard
As far as I can tell, Sean advocates both unintelligent
> and Intelligent agencies to be operating in reality. Of course the
> History of Science doesn't allow this choice. Its one or the other. We
> (= Biblical Creationism) can explain antibiotic resistance within our
> supernatural paradigm.
>
> Ray
The "history of science" has little impact on what scientists do, and
no role in allowing or prohibiting them from developing their
technical vocabulary. Only evidence has this role
wf3h
> In message <g9Yxl.14943$W06.13...@flpi148.ffdc.sbc.com>, John Harshman
> <jharshman.diespam...@pacbell.net> writes
yep my mistake...sorry.
Perplexed in Peoria
>> The No Free Lunch theorem boils down to *on average*, *over all
>> problems*, no search algorithm can perform better than random search.
>> Contrary to Dembski's claims that doesn't mean that there are no search
>> algorithm that can perform better than random search is relevant
>> domains. For example if Dembski was right the Newton-Raphson method
>> would not work.
>
> Just out of interest, how is the Designer supposed to have found the
> solutions, then? Is design just random search after all?
The Designer can solve the search problem by Brute Force. Only we
Matt Silberstein
jo...@wilkins.id.au (John S. Wilkins) in
<1ix3ryn.1vizy7at2er4nN%jo...@wilkins.id.au> wrote:
The Creator (apparently a distinct being from this Designer) did not
*find* the solutions, the Creator knew it.
>> An analogy that I offered earlier is data comprehension. It is another,
>> and possibly related, theorem of information theory that no lossless
>> compression algorithm can compress all strings (in fact the pigeonhole
>> principle implies a stronger statement should be possible). But that
>> doesn't mean that lossless data compression is impracticable in practice
>> - see the existence of ZIP, GIF, PNG etc files for counterexamples.
--
Matt Silberstein
Do something today about the Darfur Genocide
http://www.beawitness.org
http://www.darfurgenocide.org
http://www.savedarfur.org
"Darfur: A Genocide We can Stop"
John S. Wilkins
Yes, that intelligence is no better than a random search :-)
Whereas we *know* that natural selection is *way* better than a random
search.
Ergo, natural selection is way better than intelligent design.
QED
Rolf
> On Mar 23, 5:58 pm, John Harshman <jharshman.diespam...@pacbell.net>
> wrote:
>> Ray Martinez wrote:
[snip]
>>> And I wouldn't be surprised if Sean Pitman showed up here in
>>> this context. As far as I can tell, Sean advocates both
>>> unintelligent and Intelligent agencies to be operating in reality.
>>> Of course the History of Science doesn't allow this choice. Its one
>>> or the other. We (= Biblical Creationism) can explain antibiotic
>>> resistance within our supernatural paradigm.
>>
Ray, I beg on my knees: Please, please do! Explain antibiotic resistance
within your supernatural paradigm.
Don't you see that just by doing that you already will have overturned 150
years of science? Then you can take as much time as you like on the rest of
your project, you will have reached your goal: To show the world that
evolution is FALSE!
Please, don't let us (and we are no tapeworm!) down!
You said it, you said "We can." You did not say just "We might" or something
like that, you made it clear that you can, so why not go ahead and do it?
Rolf, Gnostic and "mutabilist"
John Harshman
Hard to say, because Wolpert doesn't exactly tell us what he means by
"coevolution". What he appears to mean is simply the interaction of one
part of the environment with another (either or both of which could be
organisms). Not even this is necessary, if all he needs is a changing
adaptive space. Simple climate change, for example, could do the trick.
The more important point, which Felsenstein and others have made, is
that you don't even need a changing adaptive space. All you need is a
space in which nearby points are correlated. It's pointless to average
over all possible spaces if actual spaces are a small and biased subset
of all spaces.
wf3h
>
> But the Explanatory Filter allows for the concepts seen in the
> Darwinian mechanism----that's why Dembski said what he said. Dembski
> is only denying the main positive claim of natural selection (creative
> force). And I wouldn't be surprised if Sean Pitman showed up here in
> this context. As far as I can tell, Sean advocates both unintelligent
> and Intelligent agencies to be operating in reality. Of course the
> History of Science doesn't allow this choice. Its one or the other. We
> (= Biblical Creationism) can explain antibiotic resistance within our
> supernatural paradigm.
>
for 2000 years you failed to do so. what changed? you guys had disease
as the result of 'sin' instead of bacterial contamination. NOW, all of
a sudden that science is on the scene, you got it figured out?
what a coincidence!
Kent Paul Dolan
> I stress:
> "They do on average make things worse"
> Not a word about positive, good or beneficial
> mutations, that is, the ones that an alleged
> unguided material process somehow knew to
> preserve.
Ah, I see your intellectual dishonesty generator is
in high gear again.
You couldn't be bothered actually to _read_ the
cited the article (which very much _does_ mention
beneficial mutations), that would involve due
diligence and honest effort on your part, something
any legitimate researcher does automatically, which
failure to do by you pretty much sums up your status
as a researcher: illegitimate, fraud, mountebank,
bunco artist, con man are terms that come to mind to
describe you.
You were, instead more than happy to criticize the
article's findings based only on a quoted snippet,
using your vast storehouse of ignorance and agenda
to fill in the blanks.
This intention to argue from a snippet of the whole,
in your case, from the King James Version of the
Bible and using Darwin's original works, and to
ignore the vast bulk of the pertinent literature,
the ongoing findings of science, is why your
putative and vaporware "paper" that will "overturn
evolution" is a joke before you've written the first
letter of it.
Sucks MASSIVELY to be you.
Sucks WORSE to be _known_ to be you.
xanthian.
Kent Paul Dolan
> We (= Biblical Creationism) can explain antibiotic
> resistance within our supernatural paradigm.
That would be your "the God of love hasn't given up
on trying to kill off humankind" explanation?
xanthian.
John S. Wilkins
> Ernest Major wrote:
...
> > As I understand the matter, the relevance of coevolution is that the No
> > Free Lunch theorem applies to searching *fixed* spaces - but coevolution
> > means that the spaces involved change over time.
>
> Hard to say, because Wolpert doesn't exactly tell us what he means by
> "coevolution". What he appears to mean is simply the interaction of one
> part of the environment with another (either or both of which could be
> organisms). Not even this is necessary, if all he needs is a changing
> adaptive space. Simple climate change, for example, could do the trick.
>
> The more important point, which Felsenstein and others have made, is
> that you don't even need a changing adaptive space. All you need is a
> space in which nearby points are correlated. It's pointless to average
> over all possible spaces if actual spaces are a small and biased subset
> of all spaces.
And as everyone I've ever read on natural selection from Fisher onwards
has noted, selection can only work where the adaptive landscape is
correlated - where, in the more recent terms of Kaufmann, the landscape
is smooth not rugged. A purely random landscape would be impossible to
evolve in.
John Harshman
> John Harshman <jharshman....@pacbell.net> wrote:
>
>> Ernest Major wrote:
> ...
>>> As I understand the matter, the relevance of coevolution is that the No
>>> Free Lunch theorem applies to searching *fixed* spaces - but coevolution
>>> means that the spaces involved change over time.
>> Hard to say, because Wolpert doesn't exactly tell us what he means by
>> "coevolution". What he appears to mean is simply the interaction of one
>> part of the environment with another (either or both of which could be
>> organisms). Not even this is necessary, if all he needs is a changing
>> adaptive space. Simple climate change, for example, could do the trick.
>>
>> The more important point, which Felsenstein and others have made, is
>> that you don't even need a changing adaptive space. All you need is a
>> space in which nearby points are correlated. It's pointless to average
>> over all possible spaces if actual spaces are a small and biased subset
>> of all spaces.
>
> And as everyone I've ever read on natural selection from Fisher onwards
> has noted, selection can only work where the adaptive landscape is
> correlated - where, in the more recent terms of Kaufmann, the landscape
> is smooth not rugged. A purely random landscape would be impossible to
> evolve in.
Or, to put it another way, a hill-climbing algorithm works only when
there are hills to climb.
Steven L.
> In message
> <d12a0a7a-8141-4547...@w34g2000yqm.googlegroups.com>,
> wf3h <wf...@vsswireless.net> writes
>>> On Mar 23, 8:41Â am, John Harshman <jharshman.diespam...@pacbell.net>
>>> wrote:
>>>
>>> > For reasons too complicated to explain, I am only now reading
>>> 2-year-old
>>> > issues of the Reports of the NCSE. But volume 27 #3-4 contains a great
>>> > article by Joe Felsenstein that neatly demolishes every one of William
>>> > Dembski's arguments in easy to understand language. I recommend it to
>>> > anyone, but especially Tony.
>>>
>>> > Ah, I see it's on the web too:
>>>
>>> >http://ncseweb.org/rncse/27/3-4/has-natural-selection-been-refuted-ar...
>>>
>>>
>>> BEGIN QUOTE
>>> "Dembski himself seems unable to draw this self-evident conclusion
>>> from his own argument. He acknowledges that "the development of
>>> antibiotic resistance by pathogens via the Darwinian mechanism is
>>> experimentally verified and rightly of great concern to the medical
>>> field" (Dembski 2002: 38). But by saying that he undercuts his own
Right.
What keeps closely related mutants clustered near each other in the
fitness surface is, of course, the basic laws of organic chemistry.
You're not going to get a radical random change, because the basic
chemistry of life forms consists of just a small set of important
elements (carbon, hydrogen, oxygen, etc.), though trace elements play
specific roles here and there. And the valences of carbon, hydrogen,
etc., limit the kinds of changes that can occur due to some external
driver like nuclear radiation--a peptide or a nucleic acid isn't going
to change into something radically different, because it's constrained
by the laws of chemistry.
Evolution would be much more difficult, if life forms consisted of
nearly all the elements in the Periodic Table, without carbon having its
central importance as it does in our life forms. That would make the
fitness landscape look much more random.
Which suggests that if alien life forms have evolved on some planet in
our Galaxy, their basic chemistry would also involve just a small subset
of the elements in the Periodic Table. Though not necessarily the same
elements as our life forms do.
--
Steven L.
Email: sdli...@earthlinkNOSPAM.net
Remove the NOSPAM before replying to me.
Desertphile
<wf...@vsswireless.net> wrote:
> On Mar 23, 8:18 pm, Ray Martinez <pyramid...@yahoo.com> wrote:
> > On Mar 23, 8:41 am, John Harshman <jharshman.diespam...@pacbell.net>
> > wrote:
> >
> > > For reasons too complicated to explain, I am only now reading 2-year-old
> > > issues of the Reports of the NCSE. But volume 27 #3-4 contains a great
> > > article by Joe Felsenstein that neatly demolishes every one of William
> > > Dembski's arguments in easy to understand language. I recommend it to
> > > anyone, but especially Tony.
> >
> > > Ah, I see it's on the web too:
> >
> > >http://ncseweb.org/rncse/27/3-4/has-natural-selection-been-refuted-ar...
> >
> > "Mutations
> > In the real world, mutations do not act like this. Yes, they are much
> > more likely to reduce fitness than to increase it, but many of them
> > are not lethal. I probably carry one — I have a strong aversion to
> > lettuce, which to me has a bitter mineral taste. This is probably a
> > genetic variation in one of my odorant receptor genes. It makes salad
> > bars problematic, and at sandwich counters I spend a lot of time
> > scraping the lettuce off. But it has not killed me — yet. The great
> > body of empirical information about the effects of mutation in many
> > organisms makes it clear that a great many mutations are not instantly
> > lethal. They do on average make things worse, but they do not plunge
> > us instantly back into the primordial organic soup."
> >
> > I stress:
> >
> > "They do on average make things worse"
> >
> > Not a word about positive, good or beneficial mutations, that is, the
> > ones that an alleged unguided material process somehow knew to
> > preserve.
How very odd, then, that we observe beneficial mutations happening
all the time.
> did you read his calculation of a very very minor beneficial mutation
> can, over 84 generations, come to be present in over 99% of the
> population?
>
> it doesn't take much at all.
May is too bust working on HIS PAPER to read for comprehension.
--
http://desertphile.org
Desertphile's Desert Soliloquy. WARNING: view with plenty of water
"Why aren't resurrections from the dead noteworthy?" -- Jim Rutz
Ray Martinez
Or, like I said, maybe Dembski is speaking in the context of the EF in
NFL. I understand Felsenstein's point. I wouldn't be surprised one bit
if Dembski didn't realize the full ramifications of his theory since
my own studies have convinced me that it is an all or nothing
proposition. This is why I became an immutabilist: natural selection
is impossible.
Question: is the EF a genuine model in all aspects or is the chance
component exist as a dummy concept or even straw man? If the former is
true then some type of chance phenomena would have to exist in order
for the EF to maintain its genuineness and integrity. I believe the
trend in modern day DI IDism is clear: all negative biological
production is accounted for by Darwinian mechanisms. Again, this
leaves dual antithetic causation agencies operating in reality, simply
preposterous!
> >> If Dembski's application of NFL to evolution were correct, natural
> >> selection would be impossible because nearly all mutations would be
> >> lethal. The fact that Dembski doesn't realize the implications of his
> >> argument is not in his favor.
>
> > Actually I don't have a problem with Felsenstein's point/observation.
> > It supports my view that natural selection does everything claimed or
> > does nothing at all. I am simply anticipating what Dembski might say
> > since I have seen Sean Pitman make the same point when "cornered."
>
> No, it doesn't support your view. And that may indeed be what Dembski
> might say, since he doesn't get it either, and tends to avoid responding
> to expert criticism.
>
> >>> And I wouldn't be surprised if Sean Pitman showed up here in
> >>> this context. As far as I can tell, Sean advocates both unintelligent
> >>> and Intelligent agencies to be operating in reality. Of course the
> >>> History of Science doesn't allow this choice. Its one or the other. We
> >>> (= Biblical Creationism) can explain antibiotic resistance within our
> >>> supernatural paradigm.
> >> Do you have a tapeworm again?
>
> Or some other internal parasite?- Hide quoted text -
>
> - Show quoted text -
Ray
hersheyh
wrote:
Actually we observe mutations (changes), *some* changes are more
beneficial than the variant it changed from in the environment it
finds itself in. That is, the mutation just is. The qualifier
"beneficial" is a conditional descriptor and not an inherent property
of the change.
> > did you read his calculation of a very very minor beneficial mutation
> > can, over 84 generations, come to be present in over 99% of the
> > population?
>
> > it doesn't take much at all.
>
> May is too bust working on HIS PAPER to read for comprehension.
Ray has a difference from normal human beings when it comes to
comprehension. If Billy Bob Gene Scott didn't say it, Ray can't
comprehend it. Not a trait that will be here in 84 generations.
John Harshman
How unfortunate for you, then, that natural selection is quite easy to
demonstrate in the laboratory, and only somewhat more difficult to show
in the field.
> Question: is the EF a genuine model in all aspects or is the chance
> component exist as a dummy concept or even straw man?
I don't quite understand your question. But the EF is entirely bogus, if
that answers it. It's not a model.
> If the former is
> true then some type of chance phenomena would have to exist in order
> for the EF to maintain its genuineness and integrity. I believe the
> trend in modern day DI IDism is clear: all negative biological
> production is accounted for by Darwinian mechanisms. Again, this
> leaves dual antithetic causation agencies operating in reality, simply
> preposterous!
I have no idea what you're trying to get out of the EF, which isn't
suggested to be a mechanism of anything. But we see antithetic causation
agencies in nature all the time. Lift vs. gravity; thrust vs. drag;
parasites vs. immune systems; etc.
[snip things you keep ignoring]
Ray Martinez
Let me re-phrase:
The EF sets criteria to test for things that came about by chance and
design. Is the chance component genuine or does it simply exist in
order to always conclude for design? Again, if the EF is genuine in
all aspects THEN there must exist phenomena supporting both concepts
OR the EF is simply a rigged litmus test.
> > If the former is
> > true then some type of chance phenomena would have to exist in order
> > for the EF to maintain its genuineness and integrity. I believe the
> > trend in modern day DI IDism is clear: all negative biological
> > production is accounted for by Darwinian mechanisms. Again, this
> > leaves dual antithetic causation agencies operating in reality, simply
> > preposterous!
>
> I have no idea what you're trying to get out of the EF, which isn't
> suggested to be a mechanism of anything. But we see antithetic causation
> agencies in nature all the time. Lift vs. gravity; thrust vs. drag;
> parasites vs. immune systems; etc.
>
> [snip things you keep ignoring]- Hide quoted text -
Dana Tweedy
>> We agree on this. But the point, which you seem to be unable to
>> notice, is that Dembski's interpretation of NFL doesn't. Either
>> Dembski is wrong or natural selection is impossible, though he
>> doesn't realize it.
>>
>
> Or, like I said, maybe Dembski is speaking in the context of the EF in
> NFL. I understand Felsenstein's point. I wouldn't be surprised one bit
> if Dembski didn't realize the full ramifications of his theory since
> my own studies have convinced me that it is an all or nothing
> proposition. This is why I became an immutabilist: natural selection
> is impossible.
Since natural selection is commonly and easily observed, you are wrong.
Natural selection is not only possible, it's inevetable whenever more
offspring are born than can survive.
>
> Question: is the EF a genuine model in all aspects or is the chance
> component exist as a dummy concept or even straw man?
The "chance component" is a fact.
> If the former is
> true then some type of chance phenomena would have to exist in order
> for the EF to maintain its genuineness and integrity. I believe the
> trend in modern day DI IDism is clear: all negative biological
> production is accounted for by Darwinian mechanisms. Again, this
> leaves dual antithetic causation agencies operating in reality, simply
> preposterous!
Or, your ideas are wrong. Natural selection is not antithetic to
creativity.
snip what Ray ignores
DJT
Free Lunch
<RemoveThisPref...@ix.netcom.com> wrote in talk.origins:
Cool, who did the Creator crib from?
Ray Martinez
I have flipped-flopped on the EF myself. Presently, I am undecided.
> > If the former is
> > true then some type of chance phenomena would have to exist in order
> > for the EF to maintain its genuineness and integrity. I believe the
> > trend in modern day DI IDism is clear: all negative biological
> > production is accounted for by Darwinian mechanisms. Again, this
> > leaves dual antithetic causation agencies operating in reality, simply
> > preposterous!
>
> I have no idea what you're trying to get out of the EF, which isn't
> suggested to be a mechanism of anything. But we see antithetic causation
> agencies in nature all the time. Lift vs. gravity; thrust vs. drag;
> parasites vs. immune systems; etc.
>
No, I meant natural and supernatural; guided and unguided; Intelligent
and unintelligent....
> [snip things you keep ignoring]- Hide quoted text -
>
> - Show quoted text -
And I will keep ignoring your tapeworm.
Ray
Walter Bushell
"Perplexed in Peoria" <jimme...@sbcglobal.net> wrote:
> The Designer can solve the search problem by Brute Force. Only we
> finite intellgences need to use computation and algorithms.
Omniscience implies no thinking needed.
Dana Tweedy
>> I don't quite understand your question. But the EF is entirely
>> bogus, if that answers it. It's not a model.
>>
>
> Let me re-phrase:
>
> The EF sets criteria to test for things that came about by chance and
> design.
It should be pointed out that Dembski's "explanitory filter" turns out to be
useless in practice.
> Is the chance component genuine or does it simply exist in
> order to always conclude for design?
Dembski imagines he can determine what is the product of design, and what is
"chance" (never mind that most natural processes aren't purely chance). In
practice, Dembski is only fooling himself.
> Again, if the EF is genuine in
> all aspects THEN there must exist phenomena supporting both concepts
> OR the EF is simply a rigged litmus test.
As John pointed out the 'EF' is entirely bogus. It's worthless.
DJT
John Harshman
I see no reason why we couldn't see those too. There is certainly no a
priori reason to suppose that these pairs would be impossible.
>> [snip things you keep ignoring]
>>
> And I will keep ignoring your tapeworm.
You're the one with the tapeworm. Or was it the royal "we"?
John Harshman
Nope, still no clearer. But I believe that it is indeed a rigged litmus
test; it has no application to any real-world phenomena.
Friar Broccoli
wrote:
> "John Harshman" <jharshman.diespam...@pacbell.net> wrote:
> > Perplexed in Peoria wrote:
> >> "John Harshman" <jharshman.diespam...@pacbell.net> wrote:
> >>> For reasons too complicated to explain, I am only now reading 2-year-old
> >>> issues of the Reports of the NCSE. But volume 27 #3-4 contains a great
> >>> article by Joe Felsenstein that neatly demolishes every one of William
> >>> Dembski's arguments in easy to understand language. I recommend it to
> >>> anyone, but especially Tony.
>
> >>> Ah, I see it's on the web too:
>
> >>>http://ncseweb.org/rncse/27/3-4/has-natural-selection-been-refuted-ar...
>
> >>http://www.talkreason.org/articles/jello.cfm
> >> Also nice, and after reading Felsenstein I see that it does a bit more
> >> than simply insult Dembski's clarity.
>
> > I also think that Wolpert, paradoxically, fails to make the most telling
> > argument against Dembski's appropriation of NFL, i.e. that the fitness
> > surfaces of living organisms (and genome, and proteomes) are not a
> > random sample of all conceivable fitness surfaces, but surfaces in which
> > near neighbors are typically highly correlated, i.e. those in which
> > hill-climbing algorithms are expected to do quite well. Felsenstein
> > makes this point with devastating force.
>
> Yes he does - the bit about architecture training was particularly cute.
>
> The world we live in is not encrypted. Most parts of it interact very little
> with other parts. When my family leaves home for a vacation, we have
> to make many arrangements at home concerning doors, windows, lights,
> toilets, faucets, thermostats, garbage, notifying neighbors, stopping delivery
> of newspapers, and so on. If we lived in Dembski's encrypted universe,
> this would be impossible. Every time we changed the thermostat setting,
> the windows would come unlocked and the faucets would run. Every time
> we closed a window, the newspaper delivery would resume, or a neighbor
> would forget that we were leaving. (It's worse than that, in fact. The house
> would be totally destroyed.) But, as we live in the real universe, we can
> cheerfully set family members to carrying out these different tasks without
> their worrying about each other's actions. The different parts of the house
> scarcely interact.
>
> Of course a house is a designed object, but it is not particularly hard to
> have its parts be almost independent. When architects train, they do not
> have to spend much of their time ensuring that the doors, when closed,
> will not cause the faucets to run.
>
> But Dembski might claim (and Felsenstein might concede) that the
> sheer fact that our universe is one in which that kind of modularity
> is there "for free" is itself something of an argument for the "fine-tuning"
> version of ID. So now I need to understand why it is probably the
> case that modularity is natural in most universes in the multiverse. I
> think it is, and I think I know why, but I'm not sure I can express it.
>
> Felsenstein also suggests the possibility that modularity can itself
> evolve under NS. Evolution of evolvability. That one makes me
> nervous, but it is surely possible in a multi-level selection environment.
> Trouble is, in this context, it simply re-raises Wolpert's problem
> of induction. What is it that guarantees that a mutation that makes
> one characteristic more evolvable will tend to make all characteristics
> more evolvable?
According to Lynn Helena Caporale in "Darwin In the Genome"
the opposite is the case. Some parts of the genome develop
characteristics that make THOSE regions highly prone to mutation.
She begins her discussion with the example of cone snail toxins
which are encoded by part of the genome that is highly mutation
prone.
Unfortunately the example mechanisms she described (at least
the ones I understood) did not seem to have the properties she
proposed. They were either genome wide, or they were specific
sequences that were likely to induce a higher than average
number of errors in the replication equipment.
Anyway, it seems to me that focusing mutations in areas like
bone length, and the immune system, while limiting them for
say kidney or muscle function would seem to me to be the
way to go.
>
> Architects don't spend a lot of time learning how to let in light without
> letting in rain - they put the windows on the walls rather than ceilings
> because their predecessors did. They don't spend a lot of time learning
> how to let in light without letting out heat. They glaze the windows
> because glass is implied by the word "window" in modern societies.
> And they don't spend a lot of time worrying about how to let people
> in and out without having the whole structure collapse. Nor did I
> when I recently built a shelter from brush and twine. But it collapsed
> anyways.
Ray Martinez
This comment is SO moronic. This is why we call admitted Atheist-
evolutionist John Harshman, a man who claims to have a doctorate, a
bonehead. He literally has no awareness of how stupid the comment is,
which is of course seen in the fact that he posted it.
Ray
> >> [snip things you keep ignoring]
>
> > And I will keep ignoring your tapeworm.
>
> You're the one with the tapeworm. Or was it the royal "we"?- Hide quoted text -
John Harshman
There you go again with the tapeworm. Who are "we"? Now of course if my
comment is moronic, you will be able to explain exactly what's wrong
with it. Right?
Dana Tweedy
>>> No, I meant natural and supernatural; guided and unguided;
>>> Intelligent and unintelligent....
>>
>> I see no reason why we couldn't see those too. There is certainly no
>> a
>> priori reason to suppose that these pairs would be impossible.
>>
>
> This comment is SO moronic.
Is it? How so, Ray?
> This is why we call admitted Atheist-
> evolutionist John Harshman, a man who claims to have a doctorate, a
> bonehead.
Ray, everyone knows you are jealous of John's intelligence and education.
> He literally has no awareness of how stupid the comment is,
> which is of course seen in the fact that he posted it.
So, Ray, can you explain why the comment is "stupid"? You are obviously
unable to answer it intelligently, so you resort to insults. Insult =
inability to refute.
DJT-
William Morse
> John Harshman <jharshman....@pacbell.net> wrote:
>
>> Ernest Major wrote:
> ...
>> > As I understand the matter, the relevance of coevolution is that the
>> > No Free Lunch theorem applies to searching *fixed* spaces - but
>> > coevolution means that the spaces involved change over time.
>>
>> Hard to say, because Wolpert doesn't exactly tell us what he means by
>> "coevolution". What he appears to mean is simply the interaction of one
>> part of the environment with another (either or both of which could be
>> organisms). Not even this is necessary, if all he needs is a changing
>> adaptive space. Simple climate change, for example, could do the trick.
>>
>> The more important point, which Felsenstein and others have made, is
>> that you don't even need a changing adaptive space. All you need is a
>> space in which nearby points are correlated. It's pointless to average
>> over all possible spaces if actual spaces are a small and biased subset
>> of all spaces.
>
> And as everyone I've ever read on natural selection from Fisher onwards
> has noted, selection can only work where the adaptive landscape is
> correlated - where, in the more recent terms of Kaufmann, the landscape
> is smooth not rugged. A purely random landscape would be impossible to
> evolve in.
I'm not sure I understand what mathematics would give rise to a purely
random landscape on any scale that would be relevant to natural
selection. AFAIK randomness only exists at the level of quanta, and all
our everyday experience is that of statistical mechanics. At that level
we aren't really talking about randomness but about contingent
probabilities (my best analogy is Markov chains). In short, while I agree
that a purely random landscape would be impossible to evolve in, I don't
think such a landscape can exist at the level of organisms. So
Felsenstein's space of correlated nearby points is the only kind of space
that can exist at a macro level.
Yours,
Bill Morse
John S. Wilkins
No, it implies knowing everything. The only way an omniscient being can
know everything is to have all possible scenarios internalised or
modeled. It is the same thing as a brute force method (only possibly not
temporal).
John S. Wilkins
Well that was my next point: the fitness landscape is itself the outcome
of the interactions of multiplicities of species/populations and abiotic
resources, so we can presume that in any ecosystem not in major
catastrophic failure (because, for example a bolide just hit it), the
adaptive landscape will be largely correlated already, so the FLT simply
doesn't apply - a considerable amount of the work has already been done
in any single case of selection. Another way to say this is that almost
all of any populational parameters are already subjected to strong
selection in the past.
Perplexed in Peoria
I don't see why not. Such landscapes exist at the level of computer programs.
Two machine language computer programs may differ by a single bit, but
their behavior may be dramatically different (one fit, the other completely
unfit). Yet two programs that do almost exactly the same thing may be very
different in machine code.
>> So
>> Felsenstein's space of correlated nearby points is the only kind of space
>> that can exist at a macro level.
>
> Well that was my next point: the fitness landscape is itself the outcome
> of the interactions of multiplicities of species/populations and abiotic
> resources, so we can presume that in any ecosystem not in major
> catastrophic failure (because, for example a bolide just hit it), the
> adaptive landscape will be largely correlated already, ...
I don't understand this at all. My understanding of "correlated nearby
points" is that the 'points' are points in sequence space, 'nearby' refers
to the likelihood of a mutation taking you from point A to point B,
and correlated means that nearby points have similar fitnesses (more
often than 'randomly'). So what does the health of ecosystems have
to do with it?
> so the FLT simply
> doesn't apply - a considerable amount of the work has already been done
> in any single case of selection. Another way to say this is that almost
> all of any populational parameters are already subjected to strong
> selection in the past.
And I don't understand this either. The only way the past would be relevant
would be if you claimed that we have already evolved to a region of the
landscape with gently sloping hills. I'm not sure how you would support
that claim. And it still wouldn't be relevant because Dembski's argument
is that NS could not have worked well enough in the past (without
Divine assistance). Dembski makes no particular claims about how
well NS is working now.
Rolf
> On Mar 23, 8:25 pm, John Harshman <jharshman.diespam...@pacbell.net>
> wrote:
>> Ray Martinez wrote:
>>> On Mar 23, 5:58 pm, John Harshman <jharshman.diespam...@pacbell.net>
>>> wrote:
>>>> Ray Martinez wrote:
>>>>> On Mar 23, 8:41 am, John Harshman
>>>>> <jharshman.diespam...@pacbell.net> wrote:
[SNIP]
> Or, like I said, maybe Dembski is speaking in the context of the EF in
> NFL. I understand Felsenstein's point. I wouldn't be surprised one bit
> if Dembski didn't realize the full ramifications of his theory since
> my own studies have convinced me that it is an all or nothing
> proposition. This is why I became an immutabilist: natural selection
> is impossible.
http://en.wikipedia.org/wiki/Natural_selection_(disambiguation):
Natural selection is the process by which individual organisms with
favorable traits are more likely to survive and reproduce.
Right or wrong? If wrong, why?
http://en.wikipedia.org/wiki/Natural_selection:
Natural selection is the process by which favorable heritable traits become
more common in successive generations of a population of reproducing
organisms, and unfavorable heritable traits become less common, due to
differential reproduction of genotypes. Natural selection acts on the
phenotype, or the observable characteristics of an organism, such that
individuals with favorable phenotypes are more likely to survive and
reproduce than those with less favorable phenotypes. The phenotype's genetic
basis, the genotype associated with the favorable phenotype, will increase
in frequency over the following generations.
Right or wrong? If wrong, why?
[SNIP]
Rolf
While you fail to see how that comment applies to ALL the comments you make
both here, at UD or anywhere else.
They are all moronic.
John S. Wilkins
http://www.talkdesign.org/faqs/theftovertoil/theftovertoil.html
Kent Paul Dolan
> So now I need to understand why it is probably the
> case that modularity is natural in most universes
> in the multiverse. I think it is, and I think I
> know why, but I'm not sure I can express it.
It isn't even true in this universe. The classic
example is that the calculation for where all the
balls will end up in the break of a pool table
racked triangle of balls depends _measurably_ on the
gravitational force exerted by every single massy
particle in the entire visible universe. That means,
remove one such particle at the edge of the universe
(back to the beginning of time so that the effects
of it not being there have time to propagate to the
pool table), and one would be able to detect a
visible difference in the result (where the balls
ended up) of an identical stroke of the pool cue on
the cue ball.
Since we can't _do_ that calculation, we just call
the break of a pool table setup "random", but it is
no such thing, it is in principle deterministic, as
it is too big to be subject to quantum uncertainty.
The break of a pool table ball setup, because of the
complexity of the multiple ball recoils from the
bumpers and from one another and the growth of
angular vector differences into position differences
over motion distances is a really _incredibly_
sensitive detector of forces on the pool balls,
which is what makes it suitable to prove that the
whole universe is indeed "a whole".
Thus, we know that in _some_ senses, yes, everything
in the universe affects at least _some_ cause and
effect situations.
We rather suspect that such is not the case for all
of them, or closing the garage door might kill the
lawn, and that is where your "modularity" must be
found.
So when you try to express what kind of "modularity"
this universe has, your expression of it is probably
going to be every bit as subtle as a workable, fully
acceptable description of what "life" is would have
to be, and no one has come up with one of those,
either.
> Felsenstein also suggests the possibility that
> modularity can itself evolve under NS. Evolution
> of evolvability.
That's pretty straightforward in computational
evolution, where one can choose to make the
parameters of a genetic algorithm like mutation
rate, generation to generation culling rate,
crossover rate, all evolvable values too.
Doing so makes the run of the genetic algorithm tune
itself for maximum effectiveness at the fitness part
of the ongoing evolution.
In biological evolution, tunable parameters might be
damage repair rates, types of damage repairable,
(thus tuning "mutation" by tuning how much of its
reasonably constantly created amount gets removed
before it can be expressed in a phenotype), distance
bridgeable for genes to interact to have cooperative
effects, existence in the genome of "kits" of
reusable parts like HOX sequences, proportion of the
genome that is unexpressed and so available for
reuse, and many more my ignorance of biology is too
great to know.
[I'd be particularly interested to know if
any mechanism for clearing out "junk DNA" is
confirmed, even one so simple as "it's
expensive to copy, so individuals with less
of it spend less resources on copying it
with each cell reproduction". The
incredible amount of it seen in some
organisms makes one wonder why such
considerations do not prevail.]
> That one makes me nervous, but it is surely
> possible in a multi-level selection environment.
> Trouble is, in this context, it simply re-raises
> Wolpert's problem of induction.
> What is it that guarantees that a mutation that
> makes one characteristic more evolvable will tend
> to make all characteristics more evolvable?
Why must that guarantee exist? We know by
observation that some characteristics become fixed
in the genome, so that humans share some gene
sequences with single celled organisms, where the
last common ancestor existed half a billion or more
years ago. When the "best" way to do something has
been found (where "best" is a step down a dangerous
philosophical path), evolution has no path forward,
can only make a further change by dispensing with
the need for that something instead.
xanthian.
Perplexed in Peoria
and as a perfect example of self-promotion
"John S. Wilkins" <jo...@wilkins.id.au> provided this link
> http://www.talkdesign.org/faqs/theftovertoil/theftovertoil.html
Thx, John. That helps me to better understand what Dembski
is trying to do and how he fails. But it seems to me that both
Dembski's version and the one you recommend leave out some
fairly obvious and important points about inference.
First, that there is little profit to be gained by any kind of inference
unless there is a lot of information (data) to be explained. And
you get a lot of information in either of two ways (well, actually
it is a spectrum). You can have a simple event which repeats
many times (perhaps with some variation). Or you can have
an isolated event, but with enough complexity to provide lots
of information.
Now, by stretching the usual meanings of words like 'regular'
and 'specified', you can fit both situations into the same
flowchart. For example, you can talk about the 'regularity' of
a checkerboard or about the 'regularity' of an unordered collection
of observations of galliform birds in which each male bird is
colorful, whereas each female bird is relatively drab. But it
seems odd to talk about the 'complexity' of the bird data.
Furthermore, it seems to me that both inferences to design
and inferences to natural law involve the recognition or
imposition of a 'specification' of some kind. But the difference
lies in the nature of the specification - on the causal category
which applies. We infer design when the specification
suggests final causality. And we infer natural law ('regularity')
when the specification suggests something more like efficient
or material causation. And therefore, as Wilkins and Elsberry
suggest, background information and the structure of the
specification are crucial.
Finally, sometimes different specifications complement each
other rather than compete. A simple 'regularity' explanation
of bird colors might be that sex causes coloration. But this
does not rule out the hypothesis that the rule is itself a product
of design.
Kent Paul Dolan
> And as everyone I've ever read on natural
> selection from Fisher onwards has noted, selection
> can only work where the adaptive landscape is
> correlated - where, in the more recent terms of
> Kaufmann, the landscape is smooth not rugged.
Well, no, "rugged" is okay; the genome is, after
all, changeable only in discreet steps of at
smallest one codon (though perhaps one can
interpolate by using much larger changes) so the
landscape would be a stepped one rather than a
really "smooth" one satisfying the mathematical idea
of continuity.
> A purely random landscape would be impossible to
> evolve in.
That, though, is correct.
xanthian.
Perplexed in Peoria
> [I'd be particularly interested to know if
> any mechanism for clearing out "junk DNA" is
> confirmed, even one so simple as "it's
> expensive to copy, so individuals with less
> of it spend less resources on copying it
> with each cell reproduction". The
> incredible amount of it seen in some
> organisms makes one wonder why such
> considerations do not prevail.]
It doesn't puzzle me. Consider any particular mechanism
which occasionally generates and/or eliminates junk. A
kind of transposon, for example. Two possibilities. The
mechanism either is biased toward eliminating that kind
of junk, or is biased toward producing it. If the bias is
toward elimination, then we don't see that kind of junk
and don't even notice that the mechanism exists. But
if the bias is toward production, then that kind of junk
proliferates.
So, if the kinds of junk we see all are based upon mechanisms
exerting mutation pressure leading to junk proliferation, what
holds the expansion of junk in check? It is obviously selection.
So what you should expect is a steady state. Mutation tends
to increase the amount of junk (though it occasionally decreases
it) and selection tends to decrease it. So if junk is not that
burdensome to maintain, selection will tolerate a lot of junk,
and it takes only a small mutational bias toward expansion
to generate a lot of junk.
If you want more details, I believe that Michael Lynch is
the population geneticist who has studied junk most
carefully. There were a series of papers about a decade ago
in which he explained the levels of junk in a huge variety
of organisms and organelles. For example, why is the level
of junk so different in animal mitochondria, plant mitochondria,
and chloroplasts? Lynch provides a rationalization.
Perplexed in Peoria
> And as everyone I've ever read on natural selection from Fisher onwards
> has noted, selection can only work where the adaptive landscape is
> correlated - where, in the more recent terms of Kaufmann, the landscape
> evolve in.
Not so. Trial and error works regardless of how the trials are
generated. Search eventually succeeds as long as you know what
you are looking for. The only problem is that they don't work as
quickly as you might like when performed on purely random
landscapes.
John Harshman
If I recall, genomes for which replication speed is very important
should not tolerate much junk, while it's no big deal for genomes with
slower replication. The energy costs don't seem very significant.
But what I would like explained are the huge swings between closely
related species. Why should fugu have a genome 1/10 the size of mine?
Why do frogs and ferns vary by factors of a hundred or more?
John Harshman
> John S. Wilkins wrote:
>
> > And as everyone I've ever read on natural
> > selection from Fisher onwards has noted, selection
> > can only work where the adaptive landscape is
> > correlated - where, in the more recent terms of
> > Kaufmann, the landscape is smooth not rugged.
>
> Well, no, "rugged" is okay; the genome is, after
> all, changeable only in discreet steps of at
> smallest one codon (though perhaps one can
> interpolate by using much larger changes) so the
> landscape would be a stepped one rather than a
> really "smooth" one satisfying the mathematical idea
> of continuity.
You misunderstand what "smooth" and "rugged" mean in this context. We're
not talking about strict mathematical continuity here. A smooth
landscape is simply one in which nearby points are highly correlated.
Natura facit teensy-weensy saltum.
> > A purely random landscape would be impossible to
> > evolve in.
>
> That, though, is correct.
And that's what "rugged" means, more or less. Read your Kaufmann, if you
can stand it.
Matt Silberstein
>Walter Bushell <pr...@panix.com> wrote:
>
>> In article <Wq3yl.16620$as4....@nlpi069.nbdc.sbc.com>,
>> "Perplexed in Peoria" <jimme...@sbcglobal.net> wrote:
>>
>> > The Designer can solve the search problem by Brute Force. Only we
>> > finite intellgences need to use computation and algorithms.
>>
>> Omniscience implies no thinking needed.
>
>No, it implies knowing everything. The only way an omniscient being can
>know everything is to have all possible scenarios internalised or
>modeled. It is the same thing as a brute force method (only possibly not
>temporal).
What does "all possible" mean? It is entirely reasonable to think
there is only one possible Universe for any given starting point. The
problem comes if this being has a desired end (or middle) point. Then
it requires something to figure out the start.
--
Matt Silberstein
Do something today about the Darfur Genocide
http://www.beawitness.org
http://www.darfurgenocide.org
http://www.savedarfur.org
"Darfur: A Genocide We can Stop"
John Harshman
once during the history of the universe".
hersheyh
In my view the explanatory filter of Dembski is the way pseudoscience
is done, not the way that real science works. In real science, one
does not distinguish between all testable causal explanations, pure
random explanations, and supernatural fantasy explanations, which is
what Dembski does.
The only division that *real* science makes is between *testable* and
*nontestable* explanations. The latter includes explanations that are
not currently testable because there is no currently available
evidence to test it and fantasy explanations that are inherently
untestable, like an invisible fairy did it.
The former (testable explanations) often includes explanations that
are due to probabilistic chance or the non-interaction of variables.
That is, if one can calculate the probability of something occurring
by chance under a specific set of assumptions about randomness, and
the something occurred far too frequently to fit that probability,
then the specific set of assumptions is falsified. Meaning that the
*real* explanation is, at least in part, a causal one. That is some
unknown factor is biasing the results. That leaves all the possible
causal factors one can think of and actually test as possible
scientific explanations. Including any *testable* intelligent agent
you can provide evidence for. Not including non-testable intelligent
agents you imagine exist and provide with the power to do whatever you
need for them to do.
Friar Broccoli
wrote:
> "Kent Paul Dolan" <xanth...@well.com> wrote in messagenews:gqdc1u$6f2$1...@news.albasani.net...
>
>> [I'd be particularly interested to know if
>> any mechanism for clearing out "junk DNA" is
>> confirmed, even one so simple as "it's
>> expensive to copy, so individuals with less
>> of it spend less resources on copying it
>> with each cell reproduction". The
>> incredible amount of it seen in some
>> organisms makes one wonder why such
>> considerations do not prevail.]
>
> It doesn't puzzle me. Consider any particular mechanism
> which occasionally generates and/or eliminates junk. A
> kind of transposon, for example. Two possibilities. The
> mechanism either is biased toward eliminating that kind
> of junk, or is biased toward producing it. If the bias is
> toward elimination, then we don't see that kind of junk
> and don't even notice that the mechanism exists. But
> if the bias is toward production, then that kind of junk
> proliferates.
>
> So, if the kinds of junk we see all are based upon mechanisms
> exerting mutation pressure leading to junk proliferation, what
> holds the expansion of junk in check? It is obviously selection.
I don't find this (selection as the [only] mechanism for
keeping junk in check) at all obvious.
It seems to me that pretty much any species could benefit at
least a little bit, (at least in the short term) from
eliminating most of its junk DNA.
Consequently, if some mechanism arises that preferentially
eliminates junk (I have only speculative ideas about what such
a mechanism would look like and doubt that any of those would
stand up to a pounding from HarshPerson), then pretty much any
species could wind up with a nearly junk free (or reduced)
genome.
> So what you should expect is a steady state.
If species can evolve differential junk elimination mechanisms
then there is no reason to expect to see a "steady state"
following some visible rule across a range of similarly
operating species.
> Mutation tends
> to increase the amount of junk (though it occasionally decreases
> it) and selection tends to decrease it. So if junk is not that
> burdensome to maintain, selection will tolerate a lot of junk,
> and it takes only a small mutational bias toward expansion
> to generate a lot of junk.
>
> If you want more details, I believe that Michael Lynch is
> the population geneticist who has studied junk most
> carefully. There were a series of papers about a decade ago
> in which he explained the levels of junk in a huge variety
> of organisms and organelles. For example, why is the level
> of junk so different in animal mitochondria, plant mitochondria,
> and chloroplasts? Lynch provides a rationalization.
I'm certainly interested, especially if the rationalization
consistently explains the data. I note that Harshman (in his
reply) did not mention his long standing problem with why there
is so little junk DNA in animal mitochondria. I would sure
like to get a handle on that.
Ray Martinez
Comment defines pseudoscience as that which rejects Naturalism-
Materialism. Conversely, we define pseudoscience the exact same way:
that which rejects Theism-Supernaturalism.
> In real science, one
> does not distinguish between all testable causal explanations, pure
> random explanations, and supernatural fantasy explanations, which is
> what Dembski does.
>
> The only division that *real* science makes is between *testable* and
> *nontestable* explanations.
False.
In real science there is no such thing as "non-testable" ideas or
explanations. A hallmark of pseudoscience is limitations.
Ray
> > Ray- Hide quoted text -
>
> - Show quoted text -- Hide quoted text -
Mark Isaak
> [...]
> Let me re-phrase:
>
> The EF sets criteria to test for things that came about by chance and
> design. Is the chance component genuine or does it simply exist in
> order to always conclude for design?
First, you misstate the EF; it is supposed to distinguish between chance,
natural law, and design. As such, it has no relevance to evolution,
because it never considers two or more of these in combination. It
contains the built-in assumption that all of life happened via a single
event.
--
Mark Isaak eciton (at) earthlink (dot) net
"It is certain, from experience, that the smallest grain of natural
honesty and benevolence has more effect on men's conduct, than the most
pompous views suggested by theological theories and systems." - D. Hume
John Harshman
Who is this "we" you keep referring to? Nobody defines pseudoscience in
either of these ways.
>> In real science, one
>> does not distinguish between all testable causal explanations, pure
>> random explanations, and supernatural fantasy explanations, which is
>> what Dembski does.
>>
>> The only division that *real* science makes is between *testable* and
>> *nontestable* explanations.
>
> False.
>
> In real science there is no such thing as "non-testable" ideas or
> explanations. A hallmark of pseudoscience is limitations.
Actually, a hallmark of pseudoscience is *absence* of limitations.
Anything is possible, and the pseudo-theory explains everything that
happens as well as everything that might have happened. Which means it's
unfalsifiable. Real science is falsifiable, which means that some things
that could have happened might have shown the theory to be wrong.
Limitations are essential if you are to learn anything.
And many things are untestable. String theory is currently untestable,
and may always be so. The multiverse theory, likewise. And of course God
may or may not be untestable depending on whether you are willing to
supply him with limitations.
[snip]
Ray Martinez
SNIP....
>
> > > > Or, like I said, maybe Dembski is speaking in the context of the EF in
> > > > NFL. I understand Felsenstein's point. I wouldn't be surprised one bit
> > > > if Dembski didn't realize the full ramifications of his theory since
> > > > my own studies have convinced me that it is an all or nothing
> > > > proposition. This is why I became an immutabilist: natural selection
> > > > is impossible.
>
> > > How unfortunate for you, then, that natural selection is quite easy to
> > > demonstrate in the laboratory, and only somewhat more difficult to show
> > > in the field.
>
> > > > Question: is the EF a genuine model in all aspects or is the chance
> > > > component exist as a dummy concept or even straw man?
>
> > > I don't quite understand your question. But the EF is entirely bogus, if
> > > that answers it. It's not a model.
>
> > I have flipped-flopped on the EF myself. Presently, I am undecided.
>
> In my view the explanatory filter of Dembski is the way pseudoscience
> is done, not the way that real science works. In real science, one
> does not distinguish between all testable causal explanations, pure
> random explanations, and supernatural fantasy explanations, which is
> what Dembski does.
>
> The only division that *real* science makes is between *testable* and
> *nontestable* explanations. The latter includes explanations that are
> not currently testable because there is no currently available
> evidence to test it and fantasy explanations that are inherently
> untestable, like an invisible fairy did it.
>
This comment admits to limitation, which is a hallmark of
pseudoscience. Real science has no bounds or limitations. Science has
always proven the existence of God----only Darwinists deny.
> The former (testable explanations) often includes explanations that
> are due to probabilistic chance or the non-interaction of variables.
Acceptance of the concept of "chance" to play any role in biological
production is a statement against the existence of Intelligent
causation to play any role in biological production.
Ray
> > Ray- Hide quoted text -
>
> - Show quoted text -- Hide quoted text -
redd...@bresnan.net
> On Mar 25, 9:08 am, hersheyh <hershe...@yahoo.com> wrote:
> > In my view the explanatory filter of Dembski is the way pseudoscience
> > is done, not the way that real science works.
>
> Comment defines pseudoscience as that which rejects Naturalism-
> Materialism.
Actually pseudoscience rejects the methodological naturalism that all
science uses.
> Conversely, we define pseudoscience the exact same way:
> that which rejects Theism-Supernaturalism.
Ray, what bizarre definition you use is of little consequence.
Science can only work by making use of methodological naturalism. If
you think that "supernaturalism" should be part of science, please
explain how one can test a supernatural cause.
>
> > In real science, one
> > does not distinguish between all testable causal explanations, pure
> > random explanations, and supernatural fantasy explanations, which is
> > what Dembski does.
>
> > The only division that *real* science makes is between *testable* and
> > *nontestable* explanations.
>
> False.
Sorry, but you got it wrong again, Ray.
>
> In real science there is no such thing as "non-testable" ideas or
> explanations.
How do you test the supernatural, Ray?
> A hallmark of pseudoscience is limitations.
Actually, a hallmark of pseudoscience is claims that can't be
tested. Science is limited to only such that can be observed, and
tested. Appeals to the supernatural can't be tested, which is why
they are useless to science.
Remember, I've asked you before to explain certain events via
supernatural means, and you were entirely unable to do so.
snip what Ray ignores
DJT
Ray Martinez
> On Tue, 24 Mar 2009 16:03:26 -0700, Ray Martinez wrote:
> > [...]
> > Let me re-phrase:
>
> > The EF sets criteria to test for things that came about by chance and
> > design. Is the chance component genuine or does it simply exist in
> > order to always conclude for design?
>
> First, you misstate the EF; it is supposed to distinguish between chance,
> natural law, and design.
I said the same thing a little differently.
> As such, it has no relevance to evolution,
> because it never considers two or more of these in combination.
Ambiguity.
> It
> contains the built-in assumption that all of life happened via a single
> event.
>
Really?
So you are saying the EF assumes a fact of original creation?
If so, all you are doing is identifying Creationism. You did not infer
the alleged fact from the EF itself.
Ray
redd...@bresnan.net
> On Mar 25, 9:08 am, hersheyh <hershe...@yahoo.com> wrote:
>
> > The only division that *real* science makes is between *testable* and
> > *nontestable* explanations. The latter includes explanations that are
> > not currently testable because there is no currently available
> > evidence to test it and fantasy explanations that are inherently
> > untestable, like an invisible fairy did it.
>
> This comment admits to limitation, which is a hallmark of
> pseudoscience.
No, the hallmark of pseudoscience is lack of limits. An idea that
can explain anything, really explains nothing.
> Real science has no bounds or limitations.
No, real science is bounded by what can be observed, and tested.
Without such limits, there is no way to determine which alternative is
more likely correct, and which is wrong.
> Science has
> always proven the existence of God----only Darwinists deny.
Of course, Ray defines anyone who points out his folly as a
"darwinist". A nice bit of circular reasoning. When asked to
provide an example of science proving the existence of God, Ray runs
away.
>
> > The former (testable explanations) often includes explanations that
> > are due to probabilistic chance or the non-interaction of variables.
>
> Acceptance of the concept of "chance" to play any role in biological
> production is a statement against the existence of Intelligent
> causation to play any role in biological production.
Again, Ray demonstrates his ignorance, or limited thinking ability.
Chance does not rule out "intelligent causation". Human selective
breeding is intelligent causation, but it still relies on random
mutations. Ray seems to think that "intelligent causation" is the
same thing as "supernatural causation", forgetting that humans are
intelligent (at least some of them are).
Until Ray can show some evidence of supernatural causation, the only
intelligent causation available is that provided by humans.
snip of what Ray runs away from
DJT
John Harshman
What he really means is that the EF assumes a unitary explanation.
Either all chance, or all law, or all design. It doesn't consider what
happens if a process contains some chance elements and some
deterministic elements. This is a shame, because natural selection is
exactly such a process. Thus the EF is incapable of detecting the
results of natural selection, and will in fact attribute them to design.
Ray Martinez
wrote:
SNIP....
> >>> I have flipped-flopped on the EF myself. Presently, I am undecided.
> >> In my view the explanatory filter of Dembski is the way pseudoscience
> >> is done, not the way that real science works.
>
> > Comment defines pseudoscience as that which rejects Naturalism-
> > Materialism. Conversely, we define pseudoscience the exact same way:
> > that which rejects Theism-Supernaturalism.
>
> Who is this "we" you keep referring to? Nobody defines pseudoscience in
> either of these ways.
>
Howard Hershey defined pseudoscience to be what Dembski is doing.
Dembski is a Theist who argues in favor of supernatural agents. That
was the context (review the replies). Are you saying that you disagree
with Howard?
And "we" (= Theists-supernaturalists).
And we define pseudoscience to be Darwinism-Evolution-Materialism-
Naturalism. I go even further. I define the aforementioned to be "anti-
science."
> >> In real science, one
> >> does not distinguish between all testable causal explanations, pure
> >> random explanations, and supernatural fantasy explanations, which is
> >> what Dembski does.
>
> >> The only division that *real* science makes is between *testable* and
> >> *nontestable* explanations.
>
> > False.
>
> > In real science there is no such thing as "non-testable" ideas or
> > explanations. A hallmark of pseudoscience is limitations.
>
> Actually, a hallmark of pseudoscience is *absence* of limitations.
We disagree.
Pseudoscience is acceptance of limitations, exclusionary. The word
"science" is general and all-purpose. Darwinists view the word as
sacred.
> Anything is possible, and the pseudo-theory explains everything that
> happens as well as everything that might have happened. Which means it's
> unfalsifiable.
John: evolutionary theory *claims* to explain everything.
> Real science is falsifiable, which means that some things
> that could have happened might have shown the theory to be wrong.
> Limitations are essential if you are to learn anything.
>
No one is denying falsifiability.
> And many things are untestable. String theory is currently untestable,
> and may always be so. The multiverse theory, likewise. And of course God
> may or may not be untestable depending on whether you are willing to
> supply him with limitations.
>
> [snip]- Hide quoted text -
>
> - Show quoted text -
Are you saying that String theory is not science?
Ray
wf3h
> On Mar 25, 9:08 am, hersheyh <hershe...@yahoo.com> wrote:
>
>
> >
> > In my view the explanatory filter of Dembski is the way pseudoscience
> > is done, not the way that real science works.
>
> Comment defines pseudoscience as that which rejects Naturalism-
> Materialism. Conversely, we define pseudoscience the exact same way:
> that which rejects Theism-Supernaturalism.
we know natural laws exist. they exist in chemistry and physics.
so how can SUPERNATURALISM be studied by science which studies only
NATURALISM?
ray brings up a contradiction in his thinking..
>
> > The only division that *real* science makes is between *testable* and
> > *nontestable* explanations.
>
> False.
>
> In real science there is no such thing as "non-testable" ideas or
> explanations. A hallmark of pseudoscience is limitations.
this is called 'scientism'...the idea that science can explain
everything. it can't. it can't explain
-truth
-beauty
-democracy
and other metaphysical concepts. creationism therefore is
indefensible if it includes scientism
wf3h
>>
> This comment admits to limitation, which is a hallmark of
> pseudoscience. Real science has no bounds or limitations. Science has
> always proven the existence of God----only Darwinists deny.
was einstein a scientist?
einstein was an atheist. either einstein wasn't a scientist or ray is
wrong...
now let me think....
wf3h
> On Mar 25, 1:58 pm, John Harshman <jharshman.diespam...@pacbell.net>
> wrote:
> >
> > Actually, a hallmark of pseudoscience is *absence* of limitations.
>
> We disagree.
>
> Pseudoscience is acceptance of limitations, exclusionary. The word
> "science" is general and all-purpose. Darwinists view the word as
> sacred.
can science explain the poetry of wallace stevens? or the largo from
handel's 'xerxes'? or a painting by picasso?
no. it can explain none of these. therefore science DOES have limits.
and these are easily demonstrated.
if words have meaning, then part of this meaning is exclusionary.
dark is not light. loud is not low. and science is not poetry.
>
> > Anything is possible, and the pseudo-theory explains everything that
> > happens as well as everything that might have happened. Which means it's
> > unfalsifiable.
>
> John: evolutionary theory *claims* to explain everything.
incorrect. it does not explain chemistry which is explained by quantum
physics. more garbage from ray
John Harshman
> On Mar 25, 1:58 pm, John Harshman <jharshman.diespam...@pacbell.net>
> wrote:
>
> SNIP....
>
>>>>> I have flipped-flopped on the EF myself. Presently, I am undecided.
>>>> In my view the explanatory filter of Dembski is the way pseudoscience
>>>> is done, not the way that real science works.
>>> Comment defines pseudoscience as that which rejects Naturalism-
>>> Materialism. Conversely, we define pseudoscience the exact same way:
>>> that which rejects Theism-Supernaturalism.
>> Who is this "we" you keep referring to? Nobody defines pseudoscience in
>> either of these ways.
>>
>
> Howard Hershey defined pseudoscience to be what Dembski is doing.
No he didn't. He diagnosed what Dembski is doing to be pseudoscience,
based on a prior definition (which he didn't state explicitly).
> Dembski is a Theist who argues in favor of supernatural agents. That
> was the context (review the replies). Are you saying that you disagree
> with Howard?
No. I'm saying you misunderstand Howard.
> And "we" (= Theists-supernaturalists).
Better if you speak for yourself, because I don't think mot theists
would agree with you.
> And we define pseudoscience to be Darwinism-Evolution-Materialism-
> Naturalism. I go even further. I define the aforementioned to be "anti-
> science."
That's a very silly definition. For example, it doesn't fit astrology,
homeopathy, n-rays, or koreshianity. I'm not sure you know what a
definition is. You aren't defining here. You're (like Howard)
categorizing. Unlike Howard, you don't seem to have even an implicit
definition other than "I say it's spinach and to hell with it."
>>>> In real science, one
>>>> does not distinguish between all testable causal explanations, pure
>>>> random explanations, and supernatural fantasy explanations, which is
>>>> what Dembski does.
>>>> The only division that *real* science makes is between *testable* and
>>>> *nontestable* explanations.
>>> False.
>>> In real science there is no such thing as "non-testable" ideas or
>>> explanations. A hallmark of pseudoscience is limitations.
>> Actually, a hallmark of pseudoscience is *absence* of limitations.
>
> We disagree.
>
> Pseudoscience is acceptance of limitations, exclusionary. The word
> "science" is general and all-purpose. Darwinists view the word as
> sacred.
You aren't being coherent here.
>> Anything is possible, and the pseudo-theory explains everything that
>> happens as well as everything that might have happened. Which means it's
>> unfalsifiable.
>
> John: evolutionary theory *claims* to explain everything.
No it doesn't. Wherever did you get that idea?
>> Real science is falsifiable, which means that some things
>> that could have happened might have shown the theory to be wrong.
>> Limitations are essential if you are to learn anything.
>
> No one is denying falsifiability.
?
>> And many things are untestable. String theory is currently untestable,
>> and may always be so. The multiverse theory, likewise. And of course God
>> may or may not be untestable depending on whether you are willing to
>> supply him with limitations.
> Are you saying that String theory is not science?
There are many who think so. I would say that it's a set of highly
detailed speculations that might some day give rise to science.
Something like what some IDers think ID might be, if only it were that
detailed. The thing about string theory is that it's untestable because
of physical limitations in our equipment; with better equipment, we
might be able to do the tests that are implied by the theory. Probably
not so with ID, since it's actually designed to be untestable.
John Harshman
> On Mar 25, 5:51 pm, Ray Martinez <pyramid...@yahoo.com> wrote:
>> On Mar 25, 1:58 pm, John Harshman <jharshman.diespam...@pacbell.net>
>> wrote:
>>> Actually, a hallmark of pseudoscience is *absence* of limitations.
>> We disagree.
>>
>> Pseudoscience is acceptance of limitations, exclusionary. The word
>> "science" is general and all-purpose. Darwinists view the word as
>> sacred.
>
> can science explain the poetry of wallace stevens? or the largo from
> handel's 'xerxes'? or a painting by picasso?
>
> no. it can explain none of these. therefore science DOES have limits.
> and these are easily demonstrated.
I think you confuse current limits with ultimate limits. A sufficiently
advanced neuroscience probably would be able to explain all these. If
beauty is in the eye of the beholder, close investigation of eyes should
do the trick. Of course this is far beyond our current reach. But does
that make the point invalid?
> if words have meaning, then part of this meaning is exclusionary.
> dark is not light. loud is not low. and science is not poetry.
I think that I shall never see
A poem as nice as ATP
Finish at leisure.
hersheyh
> On Mar 25, 1:58 pm, Mark Isaak <eci...@earthlink.net> wrote:
>
> > On Tue, 24 Mar 2009 16:03:26 -0700, Ray Martinez wrote:
> > > [...]
> > > Let me re-phrase:
>
> > > The EF sets criteria to test for things that came about by chance and
> > > design. Is the chance component genuine or does it simply exist in
> > > order to always conclude for design?
>
> > First, you misstate the EF; it is supposed to distinguish between chance,
> > natural law, and design.
As Dembski misuses it, the EF is supposed to distinguish between
natural law, chance, and design *in that order*. That is the only
order in which one can first *posit* (without having to supply actual
evidence) that no *possible* natural causal effects can produce the
result, then go on to reject (falsify) chance (which is the only
possibility actually *tested*), allowing one to pretend that all that
is left is supernatural causation (called design). Ignoring that he
has not *actually* ruled out natural causation, but merely assumed
that it has been ruled out.
Done in any other order the EF doesn't work. Again, in my world, the
one of *real* science, random causation is simply one of many
potentially testable causes. Rejecting random causation (which is
sometimes possible) merely means that there is some *causal* bias.
The task then is to test possible *causal* biases. If one thinks the
causal bias is some intelligent something that did something at some
time and place, it is your responsibility to test *that* specific
hypothesis by reference to empirical evidence. Rather hard to do when
the IDiot who is doing the IDing is undetectable and invisible. At
best that would merely mean you are proposing a hypothesis that cannot
currently be tested. At worst it would mean you are proposing a fairy
capable of explaining whatever you want explained.
hersheyh
> On Mar 25, 1:58 pm, John Harshman <jharshman.diespam...@pacbell.net>
> wrote:
>
> SNIP....
>
> > >>> I have flipped-flopped on the EF myself. Presently, I am undecided.
> > >> In my view the explanatory filter of Dembski is the way pseudoscience
> > >> is done, not the way that real science works.
>
> > > Comment defines pseudoscience as that which rejects Naturalism-
> > > Materialism. Conversely, we define pseudoscience the exact same way:
> > > that which rejects Theism-Supernaturalism.
Science does not *reject* theism. It is orthogonal to theism. Theism
does not work by the methodology that natural science demands.
> > Who is this "we" you keep referring to? Nobody defines pseudoscience in
> > either of these ways.
>
> Howard Hershey defined pseudoscience to be what Dembski is doing.
Pseudoscience, in fact, is exactly what Dembski is doing. He would be
a pseudoscientist, IMHO, even if he were an atheist. He is proposing
a bogus methodology in his EF as if it were a methodology that science
actually used. His filter is just plain wrong. [And not just because
he misuses even his filter by merely *asserting* that natural law has
been adequately explored before going on to chance. He also merely
*asserts* that an observed phenomenon has to be either *all* chance or
*all* causality or *all* design rather than some mixture thereof.]
There is so much wrong with the EF, as Dembski uses it, that it is
clear that he is not and never has been a real scientist.
> Dembski is a Theist who argues in favor of supernatural agents. That
> was the context (review the replies). Are you saying that you disagree
> with Howard?
>
> And "we" (= Theists-supernaturalists).
>
> And we define pseudoscience to be Darwinism-Evolution-Materialism-
> Naturalism. I go even further. I define the aforementioned to be "anti-
> science."
>
> > >> In real science, one
> > >> does not distinguish between all testable causal explanations, pure
> > >> random explanations, and supernatural fantasy explanations, which is
> > >> what Dembski does.
>
> > >> The only division that *real* science makes is between *testable* and
> > >> *nontestable* explanations.
>
> > > False.
>
> > > In real science there is no such thing as "non-testable" ideas or
> > > explanations.
In fact, as I pointed out, there are often many "non-testable" ideas
in science that do not involve invisible fairies. That is because the
necessary tools or evidence is simply not available that would allow a
workable test of an idea. In real science, we cannot merely imagine
an invisible fairy to solve all problems. I can propose what a likely
ancestor to modern whales would look like, but I cannot simply poof
the evidence (a fossil with the appropriate features into existence).
Much of science involves a *testable* "if-then" question, where,
following the "if", one states a series of assumptions, and following
the "then" a series of predictions or expectations for which evidence
can be found.
That is what happens with the chance part of Dembski's EF. He assumes
that evolution works by pure chance and then makes a perfectly valid
prediction from that assumption. When he compares his prediction with
reality, he observes a discrepancy and *should have* concluded that
evolution does not work by chance and chance alone. Note that this
does not mean that chance is not involved in evolution. Only that
evolution is not *pure* chance. There is some sort of *causal* bias
in evolution. If he had done that, and done it *before* he asserted
(without any evidence) that no possible natural causal explanation
could work, there would be no disagreement. That he then goes on to
conclude that this assertion of no *natural* causality and actual
evidence that evolution is not pure chance means that *supernatural*
design that cannot be empirically tested is a valid fall back
explanation is simply not science. It is simply "I don't have a clue,
so I'll say God did it because that makes me feel good." If he means
that *natural* design (by an actual constrained natural designer) is
the cause, then he needs to provide an empirical test of that
hypothesis.
> A hallmark of pseudoscience is limitations.
>
> > Actually, a hallmark of pseudoscience is *absence* of limitations.
>
> We disagree.
You and your tapeworm are wrong. Both of you seem to be on the
intellectual level of the tapeworm.
>
> Pseudoscience is acceptance of limitations, exclusionary. The word
> "science" is general and all-purpose. Darwinists view the word as
> sacred.
Oh, bull shit. I agree that the word "science" started out including
theology and other forms of knowledge. But when most people use the
word "science" today, they are talking about "natural science" whose
methodology for testing hypotheses requires methodological
naturalism. Otherwise, "natural science" would be meaningless and no
different from witchcraft and paganism.
> > Anything is possible, and the pseudo-theory explains everything that
> > happens as well as everything that might have happened. Which means it's
> > unfalsifiable.
>
> John: evolutionary theory *claims* to explain everything.
No it doesn't. Evolution, as used by biologists, only explains the
patterns of living organisms over time and the mechanisms that
generate those patterns. I.e., life is a pre-requisite for there to
be "biological evolution".
> > Real science is falsifiable, which means that some things
> > that could have happened might have shown the theory to be wrong.
> > Limitations are essential if you are to learn anything.
>
> No one is denying falsifiability.
You are. So is Dembski. His "design" is a fall back explanation that
is untested, untestable, and hence unfalsifiable. What is his *if-
then* hypothesis but "if I cannot explain something to my satisfaction
then I will ascribe it to design"?
>
> > And many things are untestable. String theory is currently untestable,
> > and may always be so. The multiverse theory, likewise. And of course God
> > may or may not be untestable depending on whether you are willing to
> > supply him with limitations.
>
> > [snip]- Hide quoted text -
>
> > - Show quoted text -
>
> Are you saying that String theory is not science?
It is an open hypothesis that may or may not be right. It, so far, is
consistent with the available evidence and has not been falsified.
That doesn't make it false or true. Just possible. In much the same
way, so long as it doesn't make predictions that are contradicted by
evidence, one can attribute evolution to be the way that the god or
gods form species over time.
>
> Ray
wf3h
wrote:
> wf3h wrote:
> > On Mar 25, 5:51 pm, Ray Martinez <pyramid...@yahoo.com> wrote:
> >> On Mar 25, 1:58 pm, John Harshman <jharshman.diespam...@pacbell.net>
> >> wrote:
> >>> Actually, a hallmark of pseudoscience is *absence* of limitations.
> >> We disagree.
>
> >> Pseudoscience is acceptance of limitations, exclusionary. The word
> >> "science" is general and all-purpose. Darwinists view the word as
> >> sacred.
>
> > can science explain the poetry of wallace stevens? or the largo from
> > handel's 'xerxes'? or a painting by picasso?
>
> > no. it can explain none of these. therefore science DOES have limits.
> > and these are easily demonstrated.
>
> I think you confuse current limits with ultimate limits. A sufficiently
> advanced neuroscience probably would be able to explain all these. If
> beauty is in the eye of the beholder, close investigation of eyes should
> do the trick. Of course this is far beyond our current reach. But does
> that make the point invalid?
i think ray is unclear on a distinction between the limits of science,
and scientism.
>
> > if words have meaning, then part of this meaning is exclusionary.
> > dark is not light. loud is not low. and science is not poetry.
>
> I think that I shall never see
> A poem as nice as ATP
nice touch!
John S. Wilkins
> On Wed, 25 Mar 2009 14:50:15 +1100, in talk.origins ,
> jo...@wilkins.id.au (John S. Wilkins) in
> <1ix52f2.1o9l97p1ofq1yrN%jo...@wilkins.id.au> wrote:
>
> >Walter Bushell <pr...@panix.com> wrote:
> >
> >> In article <Wq3yl.16620$as4....@nlpi069.nbdc.sbc.com>,
> >> "Perplexed in Peoria" <jimme...@sbcglobal.net> wrote:
> >>
> >> > The Designer can solve the search problem by Brute Force. Only we
> >> > finite intellgences need to use computation and algorithms.
> >>
> >> Omniscience implies no thinking needed.
> >
> >No, it implies knowing everything. The only way an omniscient being can
> >know everything is to have all possible scenarios internalised or
> >modeled. It is the same thing as a brute force method (only possibly not
> >temporal).
>
> What does "all possible" mean? It is entirely reasonable to think
> there is only one possible Universe for any given starting point. The
> problem comes if this being has a desired end (or middle) point. Then
> it requires something to figure out the start.
I don't tend to like modal realism (no reason: it just offends me) but
possible worlds are merely worlds that are stateable without
contradiction. Presumably a creator god (we are talking about theism
here) has the capacity to conceive of many possible outcomes of
creation, and chooses one of them. If omniscient, then he can conceive
of all statable worlds (and there's an infinite number of these - that's
a trivial proof). The only way he can do that is by enumeration -
stating each possible world (or modelling it, if you want to lose the
semantic aspect). Hence the omniscient creator must use a brute force
design algorithm to attain whatever utility functions he wants to
satisfy (is it sexist to use gender when talking about unknown
deities?).
--
John S. Wilkins, Philosophy, University of Sydney
scienceblogs.com/evolvingthoughts
But al be that he was a philosophre,
Yet hadde he but litel gold in cofre
John S. Wilkins
Damn. You beat me to it.
PiP, this is a problem of tractability - in order for selection to work
there have to be viable solutions within the duration of the population.
An adaptive landscape where the fitness of adjacent coordinates were not
correlated would mean that, as the correlation approaches randomness,
the chances of hitting on a fitter solution via any kind of mutation
would approach zero.
Michael Siemon
But can an omniscient creator enumerate a non-denumerable set of
possible worlds? (inquiring minds _don't_ want to know.... :-))
Kent Paul Dolan
> You misunderstand what "smooth" and "rugged" mean
> in this context. We're not talking about strict
> mathematical continuity here.
You don't get that choice. Mathematics is
wonderfully subtle. Things that are provably true of
continuous mathematical functions can be provably
false about stepwise functions, even if the smooth
spline approximating such functions has very small
approximation error.
> A smooth landscape is simply one in which nearby
> points are highly correlated.
Umm, no. Search is the subject, and the context is
very much the mathematical NFL theorem, so
evolutionists have to use the word in its
mathematical sense, they can't go off inventing new
meanings for the word and pretend their new meanings
work _in the context of_ describing the search math.
To do so risks turning reasoned discussion into
spouting of nonsense. As in ever case where a
technical vocabulary accompanies a technology, one
is _forced_ to use that technical vocabulary, and no
other, when trying to made sensible technical level
conversation about that technology's subject matter.
What you are doing is equivalent to me inventing
some completely novel definition of "clade" like
"species containing at least one gene in common" and
then trying to talk about the mathematics of
classifying and detecting the correct lineages of
nested hierarchies, using my definition of clade
rather than the one used by paleobiologists.
It wouldn't work.
xanthian.
Ray Martinez
> On Mar 25, 2:59 pm, Ray Martinez <pyramid...@yahoo.com> wrote:> On Mar 25, 9:08 am, hersheyh <hershe...@yahoo.com> wrote:
>
> snip
>
> > > The only division that *real* science makes is between *testable* and
> > > *nontestable* explanations. The latter includes explanations that are
> > > not currently testable because there is no currently available
> > > evidence to test it and fantasy explanations that are inherently
> > > untestable, like an invisible fairy did it.
>
> > This comment admits to limitation, which is a hallmark of
> > pseudoscience.
>
> No, the hallmark of pseudoscience is lack of limits. An idea that
> can explain anything, really explains nothing.
>
> > Real science has no bounds or limitations.
>
> No, real science is bounded by what can be observed, and tested.
> Without such limits, there is no way to determine which alternative is
> more likely correct, and which is wrong.
>
> > Science has
> > always proven the existence of God----only Darwinists deny.
>
> Of course, Ray defines anyone who points out his folly as a
> "darwinist". A nice bit of circular reasoning. When asked to
> provide an example of science proving the existence of God, Ray runs
> away.
>
And I have answered this request hundreds of times. Once again: Darwin
was arguing against the nine great scientific authorities that he
mentions by name on page 310 of the "Origin." Each of these men were
special creationists who accepted immutability and rejected
transmutation. These facts are basic History of Science available for
anyone to confirm. "On The Origin Of Species" (1859) is a re-
explanation theory. Darwin re-explained scientific evidence originally
uncovered by British and French natural theologians. The original
explanation supported the existence of God. Every claim seen in the
preceding sentences can be supported using scholars who are arch-
Darwinists themselves.
>
>
> > > The former (testable explanations) often includes explanations that
> > > are due to probabilistic chance or the non-interaction of variables.
>
> > Acceptance of the concept of "chance" to play any role in biological
> > production is a statement against the existence of Intelligent
> > causation to play any role in biological production.
>
> Again, Ray demonstrates his ignorance, or limited thinking ability.
> Chance does not rule out "intelligent causation".
They are antonymic concepts; enemy combatants; each says the other
does not exist----that is why they are postulated. You have shown zero
understanding of any argument based on the conceptual level.
> Human selective
> breeding is intelligent causation, but it still relies on random
> mutations.
These comments support the fact that you did not understand that the
discussion or debate presupposed the topic of this News Group; or you
have flupped up and admitted to a bait-and-switch.
> Ray seems to think that "intelligent causation" is the
> same thing as "supernatural causation", [SNIP....]
>
Of course they are.
This comment recognizes that I was arguing on-topic, which confirms my
previous observation that Dana was not.
> Until Ray can show some evidence of supernatural causation, the only
> intelligent causation available is that provided by humans.
>
> snip of what Ray runs away from
> DJT
Ray
John Harshman
> John Harshman wrote:
>
> > You misunderstand what "smooth" and "rugged" mean
> > in this context. We're not talking about strict
> > mathematical continuity here.
>
> You don't get that choice. Mathematics is
> wonderfully subtle. Things that are provably true of
> continuous mathematical functions can be provably
> false about stepwise functions, even if the smooth
> spline approximating such functions has very small
> approximation error.
If the NFL theorem doesn't apply to discrete state spaces, why are we
even talking about it?
> > A smooth landscape is simply one in which nearby
> > points are highly correlated.
>
> Umm, no. Search is the subject, and the context is
> very much the mathematical NFL theorem, so
> evolutionists have to use the word in its
> mathematical sense, they can't go off inventing new
> meanings for the word and pretend their new meanings
> work _in the context of_ describing the search math.
As I said, that's how the term is used in evolutionary biology. If NFL
requires continuous surfaces, it can't apply to evolution.
> To do so risks turning reasoned discussion into
> spouting of nonsense. As in ever case where a
> technical vocabulary accompanies a technology, one
> is _forced_ to use that technical vocabulary, and no
> other, when trying to made sensible technical level
> conversation about that technology's subject matter.
>
> What you are doing is equivalent to me inventing
> some completely novel definition of "clade" like
> "species containing at least one gene in common" and
> then trying to talk about the mathematics of
> classifying and detecting the correct lineages of
> nested hierarchies, using my definition of clade
> rather than the one used by paleobiologists.
>
> It wouldn't work.
That's true. So now what?
John S. Wilkins
Matt Silberstein
<wdNOSP...@verizonOSPAM.net> in
<J5hyl.1850$6%.966@nwrddc01.gnilink.net> wrote:
>On Wed, 25 Mar 2009 03:02:56 +1100, John S. Wilkins wrote:
>
>> John Harshman <jharshman....@pacbell.net> wrote:
>>
>>> Ernest Major wrote:
>> ...
>>> > As I understand the matter, the relevance of coevolution is that the
>>> > No Free Lunch theorem applies to searching *fixed* spaces - but
>>> > coevolution means that the spaces involved change over time.
>>>
>>> Hard to say, because Wolpert doesn't exactly tell us what he means by
>>> "coevolution". What he appears to mean is simply the interaction of one
>>> part of the environment with another (either or both of which could be
>>> organisms). Not even this is necessary, if all he needs is a changing
>>> adaptive space. Simple climate change, for example, could do the trick.
>>>
>>> The more important point, which Felsenstein and others have made, is
>>> that you don't even need a changing adaptive space. All you need is a
>>> space in which nearby points are correlated. It's pointless to average
>>> over all possible spaces if actual spaces are a small and biased subset
>>> of all spaces.
>>
>> And as everyone I've ever read on natural selection from Fisher onwards
>> has noted, selection can only work where the adaptive landscape is
>> correlated - where, in the more recent terms of Kaufmann, the landscape
>> is smooth not rugged. A purely random landscape would be impossible to
>> evolve in.
>
>I'm not sure I understand what mathematics would give rise to a purely
>random landscape on any scale that would be relevant to natural
>selection.
It is not an issue of random. The question is whether or not factors
are correlated in some way. Kauffman calls these NK landscapes and the
shape strongly depends on how on factor varies with others. If they
vary separately then you get chaos, nothing but cliffs.
[snip]
--
Matt Silberstein
Do something today about the Darfur Genocide
http://www.beawitness.org
http://www.darfurgenocide.org
http://www.savedarfur.org
"Darfur: A Genocide We can Stop"