Is Darwin's theory a tautology?
Torfi Sigurdsson
little more than basic understanding of evolutionary theory. So I rely on
you to tell me if I've completely misunderstood the concept of evolution.
(but please, no flames!)
Let's take the basic assumption "The fittest individuals survive". How is
the concept of "fitness" to be defined? There is, of course, no universal
definition; you need to be "fit" for different things in different environ-
ments. This leads to fitness defined as "ability to adapt to the environ-
ment". But what is meant by "ability to adapt to the environment?" How do
we know if one individual has managed to adapt better to his environment
than someone else? Isn't it by speculating which individual is more likely to
SURVIVE?
If "members fitness" is defined as " members adaptability to environment",
which is in turn defined as "the likelihood that a given member will survive"
Then the proposition "The fittest individuals survive" can be rephrased as
"The individual most likely to survive will survive" which is an a priori,
noncontingent, analytic, necessary truth, or simply, a tautology.
The point of my question is whether it is possible to define "fitness"
independently of "surviving", and vice versa. To me, the fact that a
given creature survives seems to me to imply that she is fit and fitness
seems to mean no more than likelihood of survival.
comments?
Torfi Sigurdsson
Jonathan Badger
fittest" -- that was Spencer, who was a "Social Darwinist" who wanted
to rid society of those he felt were "undesirable", such as the poor
and feeble minded. Unfortunately this phrase became popular and become
associated with genuine Darwinian theory, which is a shame because it
is quite an oversimplification of natural selection.
Will Stevens
...which leads to the suspicion that you may be Conducting An Experiment
on us - especially in view of the fact that what you say has often been
said and, (I would respectfully suggest) refuted before!
In a word, "The fittest individuals survive" is not an adequate statement
of the theory of evolution by natural selection. Consider this rather
more detailed account:
1. In most species, far more young are produced than can possibly
survive to reach reproductive maturity and to produce young of their own.
2. Which individuals survive and reproduce is, to some extent,
causally related to particular characteristics which they possess.
3. Some of these characteristics are heritable, and so tend to be
passed on from an individual to its progeny.
4. So there's a tendency, over time, for such characteristics to
spread through a population.
Where's the tautology?
Presumably, you might say "In statement 2." In which case, consider the
following analogous statements: "Which businesses survive in a recession
is causally related to the management methods used" and "Which marriages
last for more than 10 years is causally related to how the partners treat
each other." Surely you wouldn't argue that _these_ statements are
tautologies also? - though they are certainly general, and need to be
followed up by work to determine which characteristics are involved and
how powerful their effects are.
Incidentally, in the case of evolution, much work _has_ been done with
many species to identify the characteristics related to survival and to
determine the magnitude of their effects. The results show that there's
no doubt that evolution by natural selection actually happens.
A further point which might interest you as a psychologist: quite often,
when discussing evolution, people give serious attention and respect to
forms of argument which they would instantly dismiss in most other
contexts. It's interesting to speculate why. I'd suggest that it's
indicative of an aversion to the idea of evolution and of a desire not to
accept it.
-=Will Stevens=- wste...@cix.compulink.co.uk
Onar Aam
Survival of the fittest is indeed a tautolgy. (not that there is
anything wrong in that. All internally consistent conceptual
frameworks are tautologies.) But the statement needn't have to be
a tautology. Fitness can take on a definition which intrinsically
separates it from survival. This is done by Ben Goertzel in his
"The Evolving Mind". He uses fitness as in "fit like a glove".
In other words, a social/ecological concept. The underlying
assumption here is that ecologies are governed by emergent
differentiation. In ecologies emergent interrelations are formed
(e.g. different niches) and Goertzel's idea is that fitness means
that these structures don't "crash" or intersect. Species will
tend to avoid each others niches. Thus, his contention is that
those species that fit into an ecological structure will tend to
survive.
Onar.
Richard Brodie
interesting question, one which has been asked repeatedly in the century or so
since Darwin published and since Spencer popularized that phrase.
The short answer is that "survival of the fittest" is more a PR slogan than a
scientific theory. Darwinism is *really* about *replication* of the fittest. And
not individuals, but genes (in biology) or memes (in culture).
To say that the world is populated by a process of the fittest replicators
proliferating is not a tautology; it's a physical model that seems to fit the
historical data. Just as one could imagine a universe with no gravity, one could
conceive of another universe in which replication was impossible and evolution
occurred by substances mixing or transforming in some way.
Richard Brodie - RBr...@msn.com - author of the books:
"Virus of the Mind: The New Science of the Meme" (Integral Press, 1996)
"Getting Past OK: A Straightforward Guide to Having a Fantastic Life" (Warner, 1993)
CEO, Brodie Technology Group, Inc., Seattle, a technology consulting firm.
Arlin Stoltzfus
>The point of my question is whether it is possible to define "fitness"
>independently of "surviving", and vice versa. To me, the fact that a
>given creature survives seems to me to imply that she is fit and
> fitness seems to mean no more than likelihood of survival.
>
There are many reasons why evolutionary theory as a whole is not
tautologous. For instance, there is a large body of modern evolutionary
theory about how 'the neutral' (as opposed to 'the fit') may survive.
Even the apparent tautology that the fit survive because the fit survive
is not a tautology, because the fit (1) is different from the fit (2).
In one case, we are talking about individuals that may thrive and
reproduce (or fail to do so) and in the second case the "survival" is a
more long-term evolutionary "survival", not really the same thing at
all. Just to give an example, if the differences in fitness are not
hereditable, then we can still have natural selection favoring those who
survive (everyday sense) in the short term, but this selection does not
influence their chance to survive (evolutionary sense) in the long term.
Arlin
Arlin Stoltzfus (ar...@ac.dal.ca)
Department of Biochemistry
Dalhousie University
Halifax, NS B3H 4H7 CANADA
ph. 902-494-3569; fax 902-494-3569
Phil Nicholls
>Survival of the fittest is indeed a tautolgy. (not that there is
>anything wrong in that.
A tautology, as I recall from my formal logic class, a tautology is a
compound statement that is always true regardless of the truth value
of it's component statements.
Survival of the fittest is not a compound statement and therefore is
not a tautology.
Phil Nicholls pn...@globalone.net
"To ask a question you must first know most of the answer"
-Robert Sheckley
Herb Huston
The expression often used by Mr. Herbert Spencer, of
the Survival of the Fittest, is more accurate, and is
sometimes equally convenient.
-- Charles Darwin, _The Origin of Species_.
This survival of the fittest which I have here sought
to express in mechanical terms, is that which Mr. Dar-
win has called "natural selection, or the preservation
of favoured races in the struggle for life.
-- Herbert Spencer, _Principles of Biology_.
--
-- Herb Huston
-- hus...@access.digex.net
Onar Aam
>To say that the world is populated by a process of the fittest replicators
>proliferating is not a tautology
Oh, but it is. You still haven't avoided a tautological definition of
fitness. Fitness still means survival in the sense of replicative success.
Onar.
Wirt...@aol.com
>>Survival of the fittest is indeed a tautolgy. (not that there is
>>anything wrong in that.
>A tautology, as I recall from my formal logic class, a tautology is a
>compound statement that is always true regardless of the truth value
>of it's component statements.
>Survival of the fittest is not a compound statement and therefore is
>not a tautology.
Phil is exactly correct -- but only half exactly correct :-). There are two
common, formal definitions of the word tautology, in addition to the informal
one that everyone in biology seems to presume is true, that is, that a
tautological statement is indicative of circular reasoning.
As Phil writes, in formal logic, a tautology is a compound statement such
that the universe of all possible outcomes is covered and the answer is
always yes. An example of such a statement is the common question: "Well, do
you want to go or don't you?" The only correct answer is "yes", but your
questioner rarely finds that answer amusing.
But in English grammar however, a tautology is quite differently defined and
is merely the needless repetition of words or ideas, as in the phrases:
"pernicious harm" or "redundant repetition."
"Survival of the Fittest" is a phrase that comes much closer to the second
definition than the first, but it doesn't truly qualify as a tautology in
either sense. Nor is there any implication of the common presumption of
circular reasoning in either formal definition.
As any number of people have pointed out, Darwinian evolutionary theory is
not a scientific theory in the classical sense. Rather it is hardly much more
than a recipe for an iterated optimization algorithm, where each cycle
through the loop increasingly sieves the most competitive, most appropriate
("fit") exemplars from the lesser-fit variants that currently exist within
the inevitable Malthusian populational excess. The surviving phenotypes
become the parents for the succeeding generation and the loop proceeds one
more iterative cycle, in indefinite repetition, constantly culling the least
appropriate phenotypes from the interbreeding population. While the process
is indeed iterative, there is nothing particularly any more "tautological"
about the process than there is any other form of commonly iterated
optimization [indeed, the phrase "iterated optimization" could easily be
described as a tautology (in the grammatical sense), because there can be no
form of algorithmic optimization other than iterative -- with the exception
of perhaps: "Step 1. Choose the correct answer and quit"].
Wirt Atmar
Jonathan Badger
>Jonathan Badger <bad...@phylo.life.uiuc.edu> wrote:
>}The first thing to remember is Darwin never said "survival of the
>}fittest" -- that was Spencer, who was a "Social Darwinist" who wanted
>}to rid society of those he felt were "undesirable", such as the poor
>}and feeble minded. Unfortunately this phrase became popular and become
>}associated with genuine Darwinian theory, which is a shame because it
>}is quite an oversimplification of natural selection.
> The expression often used by Mr. Herbert Spencer, of
> the Survival of the Fittest, is more accurate, and is
> sometimes equally convenient.
>-- Charles Darwin, _The Origin of Species_.
Interesting -- What edition of the _Origin_, do you know? It's not in
my copy of the first, but I know Darwin added a lot of additional
comments in later editions. I'm surprised that Darwin would lend
credence to Spencer's ideas, though.
Tim Ikeda
It need not. One can define fitness as the flux through a metabolic
pathway for a growth-limiting metabolite. (see papers out of
DE Dykhuizen's and DL Hartl's labs -- also Dykhuizen, DE "Chemostats
used for studying natural selection and adaptive evolution." Methods
in Enzymology, 1993, 224:613-31). Note also that the phrase "survival
of the fittest" would have no relevance to evolution if "fitness"
characteristics were not inherited. That phrase is but one part of
Darwin's theory.
Regards, Tim Ikeda (ti...@mendel.berkeley.edu)
Onar Aam
>Webster's states that a tautology is "needless repetition of an idea in a
>different word, phrase, or sentence; redundancy." In life history theory
>fitness is defined in terms of survival AND reproductive success and
>there is often an implied tradeoff. Thus, "survival of the fitness" is NOT
>a tautology.
...and...
S.Labonne writes:
>Wrong. 1) Fitness has at least as much to do with reproductive
>success as with survival.
Obviously when we speak of survival in an evolutionary context we mean
inter-generational survival, i.e. reproductive success. It is
meaningless to speak of survival within lifespans because life is
100% lethal. All organisms die sooner or later and exactly how long they
live is uninteresting. All that matters is whether they manage to produce
offsprings before they die. Thus, survival of the fittest is still a
tautology. Fitness as it is used today is a redundant way of saying
reproductive success. HOWEVER, it doesn't have to be this way.
Tim Ikeda writes:
>It need not. One can define fitness as the flux through a metabolic
>pathway for a growth-limiting metabolite.
And, as I originally pointed out, Ben Goertzel defines fitness as in
the term "fits like a glove". Thus, he uses it as an ecological/social
concept. This use of the word is not tautological because fitness as
he defines it has nothing to do with reproductive success. Then the
statement "survival of the fittest" becomes highly interesting and
non-obvious. It implies a correlation between reproductive success
and a species' behavior relative to the ecological context it finds
itself in. *This* is not a tautology.
Onar.
Onar Aam
>So, it's an ITERATIVE system, and NOT a cycle. The "Tautology" only
>arises if we accept the first illustration, which is clearly bogus.
>
>As another example, think of the generation of the Mandelbrot set:
> Z=Z*Z+c
>This is a similar tautology, unless we clearly label the left-most
>Z as Z' (Z prime), meaning that the new Z is different from the
>old Z.
>
>In evolution, the new Fittest (F') is different from the old
>Fittest (F) in exactly the same way. This is a fundamental concept
>in SotF and evolution, and one which, unfortunately, gets
>overlooked when we get too caught up in word-games.
I agree that reproduction is an iterative system. I don't agree that
it is not circular, because iterative systems are fundamentally circular,
not just semantically. It is almost obvious that RE-production is
circular. If the offspring did not have an element of the same as the
original then it wouldn't be re-production. Although the offspring may
be slightly modified the fundamental circularity of reproduction is
always preserved. (if reproductivity is lost then the iteration stops)
This is completely analogous to the iterative function z' = z^2 + c.
Even though the particular values of the function varies from iteration
to iteration the function remains the same. That is, after 1,000,000
iterations the function is still z' = z^2 + c. Hence the _structure_
of the system is fundamentally circular. So it is not a matter of
word games. The circularity is real.
Onar.
Shane McKee
>Shane McKee wrote:
>>So, it's an ITERATIVE system, and NOT a cycle. The "Tautology" only
>>arises if we accept the first illustration [a circularity], which
>>is clearly bogus.
>I agree that reproduction is an iterative system. I don't agree that
>it is not circular, because iterative systems are fundamentally circular,
>not just semantically. It is almost obvious that RE-production is
>circular. If the offspring did not have an element of the same as the
>original then it wouldn't be re-production. Although the offspring may
>be slightly modified the fundamental circularity of reproduction is
>always preserved.
Yes, _but_ each time it goes through the machine, it is a new run
of the loop. The phenomenon of circularity is really just a
semantic convention to save us having to run the whole scenario
over in our heads many times. Each reproduction, each generation is
a new event. Each iteration of the Mandelbrot function is a new
event. That's how change occurs. If it was a true circularity,
there could be no change.
In my first post I mentioned that the 'Fittest' that comes out of
the machine is different from the 'Fittest' that goes into the
machine. If it were not so, then the tautology would hold. Even if
the cycle is structurally the same, the input and output change.
No tautology there.
(if reproductivity is lost then the iteration stops)
>This is completely analogous to the iterative function z' = z^2 + c.
>Even though the particular values of the function varies from iteration
>to iteration the function remains the same. That is, after 1,000,000
>iterations the function is still z' = z^2 + c. Hence the _structure_
>of the system is fundamentally circular. So it is not a matter of
>word games. The circularity is real.
Maybe I hadn't explained it well enough... the 'tautology' only
exists if the zs & z's necessarily keep the same values. And they
don't.
Dr Shane McKee. Royal Hospitals, Belfast, Northern Ireland UK
--[new .sig under construction. Delays may be expectorated]--
(Could someone please get the IRA to stop putting guns to our
heads so that we can have a peace process? Thanks in advance.)
Dr. David Rosen
>of the machine that change.
>
The machinery has changed. The reproductive system, and accessories,
sometimes work slightly differently. After a few runs, The ACCUMULATED effect
is a new machine.
dro...@inxs.chem.duke.edu
Luke Kaven
>Survival of the fittest is indeed a tautolgy. (not that there is
>anything wrong in that. All internally consistent conceptual
>frameworks are tautologies.) But the statement needn't have to be
...
>Onar.
A tautology is only something that is _deductively_ true, and therefore
it is not apppropriate for an inductive maxim such as "the fittest
shall survive" or some such statement. Elliot Sober explores this
a bit in a chapter in _The Philosophy of Biology_. There is to me
a disturbing circularity in the statement-type in question, but
the statement has never been used as a definition as far as I am
aware.
Luke Kaven
Center for Advanced Biotechnology and Medicine
Rutgers University
Eystein Markusson
<sh...@reservoir.win-uk.net> wrote:
As I see it, the following taken from an aticle by Arthur L. Caplan
(1985. The nature of Darwinian explanations: Is Darwinian evolutionary
theory scientific? In What Darwin began. Godfrey (ed). Ally and Bacon)
show that it is not a tautology:
The Argument for a Struggle for Existence
Empirically verified principles-
(P1) Principle of Reproduction: Nearly all organisms possess both
the capacity and the drive to reproduce themselves at rapid rates.
(P2) Principle of Dependence: All organisms depend on
naturalresuorces such as air, water, and energy for life.
To these principles Darwin adds three factual observations:
Factual observations-
(F1) There are a limited number of places on the earth where
natural resources exist.
(F2) The locations where natural resources can be found fluctuate
over time.
(F3) A good deal of variation exists in the form and behavior of
organisms.
Darwin then aruges that the facts he has cited, when combined with the
principles he has discerned to be operative in nature, lead directly to
two important conclusions:
Logically necessary conclusions derived from principles and facts-
(C1) There must be obtain a competition or a "struggle for
existence" among organisms for natural resources.
(C2) The struggle for existence is a possible source for the
patterns of quantity and variation found among the organisms dwelling on
the earth"
And further:
The Argument for Natural selection
Empirically verified principles-
(P1) Certain traits and behaviors will allow organisms that possess
them to obtain more resources than organisms that do not.
(P2) Organisms tend to reproduce similar organisms.
(P3) Organisms possessing advantageous traits and behaviors will
tend to have higher survival and, thus, reproductive success than
organisms possessing less advantageous traits and behaviors.
Factual observations-
(F1) A struggle for existence exists in nature.
(F2) A great deal of variation exists in the traits and behaviors
possessed by organisms.
(F3) Some traits and behaviors, when assessed relative to other
traits and behaviors, can be classified as advantageous, neutral, or
disadvatageous in securing scarce natural resources.
Logically necessary conclusion deducible for principles and facts-
(C1) There will be a natural selection among the variegated traits
and behaviors of organisms tending to favor an increase in organisms
possessing the most advantageous traits and behaviors relative to those
posessed by other organisms"
Eystein Markusson
Department of Ecology/Zoology tlf: +47 776 44 454
Institute of Biology and Geology email: eyst...@ibg.uit.no
University of Tromsoe
Norway
Onar Aam
But the machine is always capable of reproduction. This is wherein the
tautology lies. Only the machines that are capable of reproduction
will reproduce.
Onar.
Shane McKee
(I'm preserving this for continuity's sake...)
>>> Although the offspring may
>>>be slightly modified the fundamental circularity of reproduction is
>>>always preserved.
>>
>>Yes, _but_ each time it goes through the machine, it is a new run
>>of the loop.
>
>Yes, but it is the _same_ machine that it goes through. Hence there must
>be a circular motion of some kind.
>> The phenomenon of circularity is really just a
>>semantic convention to save us having to run the whole scenario
>>over in our heads many times.
>
>Nope, if the machine had changed then I would agree, but it is only the values
>of the machine that change.
Well, this is getting a little academic, because I don't think
we're essentially disagreeing in this exchange. I's a philosophical
point, but, I'm going to stick to my position that the iterative
system is not a truly circular one. You need to remember that
there is an extra dimension to this process, which is [Time], over
which the 'circularity' occurs. If you include [Time] in the
process (or any progressive dimensional constraint), the picture
is helical, and not circular.
Of course, if you look at it end-on, with [T] eliminated, it will
look circular, but you lose the fundamental nature of the
iterative process - in other words, you're losing an important part
of the machine.
Note that I am not disagreeing with you - it's just that someone
posted that SotF was a tautology, and I do not agree with this,
since our iterative process is one-way. I see a tautology (in the
sense meant by the original poster) as a process that you can turn
around, and it would still be the same. You can do that with a=b
-> b=a, but you can't do it with
...-->F1-->(survival)-->(reproduction)-->F2-->...
I know that you know this, but I'm not altogether sure that the
original poster, among others, appreciates the subtlety of the
distinction between the two scenarios.
Shane
Dr Shane McKee. Royal Hospitals, Belfast, Northern Ireland UK
-----[Sig-works in progress. Please use Near-side verge]-----
Onar Aam
>A tautology is only something that is _deductively_ true, and therefore
>it is not apppropriate for an inductive maxim such as "the fittest
>shall survive" or some such statement. Elliot Sober explores this
>a bit in a chapter in _The Philosophy of Biology_.
How hard can this be. How do we measure fitness? With survival, of
course. Thus, we are claiming that the fittest survive and at the
same time measure fitness with survival. Smells like tautology to me.
Onar.
S. LaBonne
>> The machinery has changed. The reproductive system, and accessories,
>>sometimes work slightly differently. After a few runs, The ACCUMULATED effect
>>is a new machine.
>
>But the machine is always capable of reproduction. This is wherein the
>tautology lies. Only the machines that are capable of reproduction
>will reproduce.
Duh, yup, _that_ is a tautology, all right. But what does it have to
do with Darwinism, which deals with competition AMONG the "machines
that are capable of reproduction"? Your point, if any, seems to have
evaporated.
--
Opinions are mine alone; I never met a university with opinions!
Steve LaBonne ********************* (labo...@cnsunix.albany.edu)
"It can never be satisfied, the mind, never." - Wallace Stevens
Onar Aam
SHane McKee writes:
>Well, this is getting a little academic, because I don't think
>we're essentially disagreeing in this exchange. I's a philosophical
>point, but, I'm going to stick to my position that the iterative
>system is not a truly circular one. You need to remember that
>there is an extra dimension to this process, which is [Time], over
>which the 'circularity' occurs. If you include [Time] in the
>process (or any progressive dimensional constraint), the picture
>is helical, and not circular.
>Of course, if you look at it end-on, with [T] eliminated, it will
>look circular, but you lose the fundamental nature of the
>iterative process - in other words, you're losing an important part
>of the machine.
I see that we are very close on our conception of these processes.
Given your linear concept of time I follow your helical perspective
completely. But in my view it is equally possible to base the argument
on circular time in which case there is no motion along a time axis.
A classic example of a phenomenon that is best described in circular
time is self-fulfilling prophecies. Here an alleged future event (e.g.
a crash in the stock market) becomes the very cause of that event.
This _can_ be explained linearly, but that explanation is not very
enlightening. The same I think is true for all iterative/circular/
feedback processes. In a world where there are both circular and
non-circular processes then both the linear and the circular concept
of time must be applied simultaniously. The most obvious way to do
this is to view organisms as time-machines. Our best indication that
time is passing is the continuous change and/or degeneration of the
surroundings. This change is then a rough measure of linear time.
Organisms, however, don't change or do so very slowly. As such we
may see them as systems that slow down linear time. An organism
approximately stands still in time. A million years from now many
mountains will have considerably changed and eroded. Even so we
will probably find animals that look the same as they do today. It
is as if time has stood still for them. It is in the light of this
that we have to understand natural selection. Natural selection
tends to preserve those systems for which time stands still. My
point here is that the confusion concerning tautology etc. is not
due to some serious flaw in the theory of natural selection but to
the very circular nature of the system at hand.
>I see a tautology (in the
>sense meant by the original poster) as a process that you can turn
>around, and it would still be the same. You can do that with a=b
>-> b=a, but you can't do it with
>...-->F1-->(survival)-->(reproduction)-->F2-->...
Oh, but you can. The key here is how we _measure_ fitness. Start with
the statement "survival of the fittest". This basically states that
survival = fittest
But how do we measure fitness? With survival of course! Then we have
that
fitness = survival
Hence, a tautology in the sense of the original poster. As long as we
don't treat fitness any different from survival then the statement is
a tautology.
Onar.
Peter H. Weis
Content-Type: text/plain
In article <4881a9$6...@nntp5.u.washington.edu>, Shane McKee <sh...@reservoir.win-uk.net> says:
>
>Onar Aam writes...
>
>>Shane McKee wrote:
>>>So, it's an ITERATIVE system, and NOT a cycle.
(snip)
>
>>I agree that reproduction is an iterative system.
(snip)
...iterative systems are fundamentally circular,
>>not just semantically. It is almost obvious that RE-production is
>>circular. If the offspring did not have an element of the same as the
>>original then it wouldn't be re-production. Although the offspring may
>>be slightly modified the fundamental circularity of reproduction is
>>always preserved.
>
Each iteration of the Mandelbrot function is a new
>event. That's how change occurs. If it was a true circularity,
>there could be no change.
>
>(snip)
>
>>This is completely analogous to the iterative function z' = z^2 + c.
>>Even though the particular values of the function varies from iteration
>>to iteration the function remains the same. That is, after 1,000,000
>>iterations the function is still z' = z^2 + c. Hence the _structure_
>>of the system is fundamentally circular. So it is not a matter of
>>word games. The circularity is real.
>
>Maybe I hadn't explained it well enough... the 'tautology' only
>exists if the zs & z's necessarily keep the same values. And they
>don't.
>(snip)
>
>
The Mandelbrot algorithm is a very useful and vivid
illustration of how the *Increasing Complexity* of the
Mandelbrot set is generated by one continuously *changing*
factor (Z), and one invariant or *stable* factor (C). As such
it illustrates very nicely how the Increasing Complexity of
gestation (as distinct from the more general term
*reproduction*) and growth are also generated by the nested
hierarchies of the well known physical, chemical, and
biological (metabolic, genetic, morphological) *Change* and
*Stability* factors. That the images of the Mandelbrot set are
also *organic* in a hauntingly beautiful way, and have the
same *self-similarity* so familiar in the biological
realm is therefore quite natural. Sometimes, not often but
sometimes, its seems for all the world that one or the other
item in the Mandelbrot has given birth to a whole raft of
offspring, and is ready to swim away at any moment. Other than
this, the Mandelbrot algorithm is not very useful since its
iteration is exponential, whereas gestation and growth are
geometrical in nature. However, if the Mandelbrot algorithm is
re-written so that the iteration of Z becomes geometric
instead of exponential, then we would have a universal
algorithm for all gestation and growth. But this lands us
squarely in the generally uncharted territory of *Increasing
Complexity* - and at severe odds with the 2nd law of
thermodynamics. However, it also lands us at a point where it
suddenly becomes clear that the Increasing Complexity of the
evolution of the universe, of Life and of Growth are all
driven by the same fundamental dynamics - particularly since
the Dynamics of Change & Stability arise directly, explicitly
and very beautifully out of the 1st law of thermodynamics. And
this lands us right at the root-core of our universe itself -
in the nature of *Energy* itself.
Although severely restricted by the text based limitations
of this medium, this little diagram conveys the essence of the
dynamics of all evolution.
ANENTROPY
! (+)/(-)
! Y / Y
C ! X / T
! E / I
H ! L / X
! P / E
A ! M / L
! O / M ENTROPY
N ! C / O
! / C
G ! R /
! C / R
E ! N / C
!I / E
! / D
ENERGY 0----------------------------
S T A B I L I T Y
THE DYNAMICS OF INCREASING COMPLEXITY AND DECAY
*Increasing Complexity* is clearly directional along the
*upwards' direction of the line of increasingly greater
Complexity - be this of gestation, growth or the whole process
of evolution (the *positive* or *constructive* side of the
diagonal line); whereas decay is equally directional in the
opposite or *downward* direction - (the *negative* or
*reductive* side of the diagonal line).
This resolves our conflict with the old 2nd law - very
simply, very elegantly, and very beautifully - and we can now
see that the old 2nd law, while far from wrong, was just too
short; too short by half.
That *Energy* itself (if in the forms of ambient and food
energy) is the generating and governing factor of the
evolution of Life, of reproduction, of growth, of reproductive
success in the Darwinian sense, as well as of decline and
death - although long tacitly understood - is now becoming a
rapidly clearing realisa- tion among many scholars.
And instead of the old and familiar Darwinian theory of
evolution we now find that we have the foundation and
framework for a brand new and truly *Universal* theory of
evolution on our hands here - a theory which accounts for the
evolution of the universe and of Life alike in the same
breath, so to speak - and does so, as it should, very simply,
very elegantly, and very, very beautifully.
And as Mitchel Feigenbaum has shown us with his "Chaos
Dynamics", the reproduction of populations is profoundly
*chaotic* in nature, and rises or falls according to available
resources, the weather, climate, predation, and other
profoundly chaotic factors. That deterministic *Chaos* plays a
fundamental role in evolution is another currently emergent
realisation among scholars. In this context, it now turns out
that Chaos and Order are a subset of Change & Stability
dynamics since Chaos is extreme Change, and the *attractors*
or *strange attractors* of Chaos (the irreducible Order in
Chaos) are simply Stability factors. Since the quintessence of
Chaos is that it never repeats itself, we also now have the
root-core of all diversity, of the unique indivi- duality of
all living things, and thus of the driving force of all
evolution.
To summarize: Besides unifying the dynamics of Change &
Stability, as well as of Chaos & Order as a subset of Change &
Stability Dynamics, this little diagram contains the 1st, 2nd
and 3rd laws of thermodynamics, derives the 2nd and 3rd laws
of TD from the 1st; completes the *old* and incomplete 2nd
law, gives us the Dynamics of Increasing Complexity, and thus,
the universal
foundation and dynamics of the evolution of the universe and
of Life alike, gives us the dynamics of gestation and growth,
and among a myriad of other profound insights, gives us the
dynamics of instinctual, social, psychological, cognitive and
technological evolution. It also makes our universe
*constructively* dynamic, as a single glance at the firmament,
and the Earth under our feet, so readily confirms. A
constructively dynamic universe which generates increasing
complexity (every star in the universe is a blazing factory of
the increasing complexity of the elements) is also a
teleological universe which contains a plan and a purpose - as
so faithfully reflected in the teleological nature of all
gestation and growth (if the *energy* point of the diagram is
taken to represent the first living thing on the Earth it
describes the evolution of Life, and when taken to represent a
fertile ovum, it gives us the dynamics of gestation and
growth). Clearly, this theory goes far deeper than
Darwinian theories but leaves Darwin's insights of common
descent and natural selection intact; it just goes a lot
deeper - to the ultimate physical origin and driving factors
of all evolutionary Change, all Stability (both morphological
and physiological), Increasing Complexity, the emergence of
new and higher levels of Complexity in evolution, as well as
the new Dynamics of Maturity (Stability). It also gives us a
completely new, and powerfully cohesive insight into the
dynamics of evolution which solves many of the passionate
arguments now swirling around this subject. It also makes the
evolution of Life - and biology in general - as 'hard' a
discipline as math and physics. On the other hand, and
unfortunately perhaps, this also makes a course in theoretical
mathematics a basic requirement in Biology.
This framework is currently being used as the foundation
for a new theory of the evolution of Life by some scholars at
the University of Zuerich. This is fine with me, and I had
intended the original posting in this newsgroup ("A New
Universal Theory Of Evolution") for this purpose - as long as
credit is given for my original insights and framework for
such a theory, and the several other theories which are
contained in these insights.
Of necessity, this posting is a bare-bones sketch which
skips over many of the beautiful details of these dynamics,
among them the *Strata of Stability*, the directly related
*Strata Of Progressive Complexity*, the first approximation of
the complete 2nd law of TD, as well as a lot of other
beautiful stuff. But all of this follows naturally out of the
little diagram, and can be worked out by anyone. With the
caution that a large background of essential detail lies
behind this sketchy outline, I will gladly discuss this in
greater detail and look forward to any comments.
peter;
--*-*-*- Next Section -*-*-*--
Shane McKee
>Organisms, however, don't change or do so very slowly. As such we
>may see them as systems that slow down linear time. An organism
>approximately stands still in time. A million years from now many
>mountains will have considerably changed and eroded. Even so we
>will probably find animals that look the same as they do today.
Yes, but the point is that they are *not* the *same* organisms. The
code for those organisms has passed through the iterative mill,
been jumbled up and rearranged. The individual organisms were
born, got old (or maybe not!) and died. Time does not stand still
for them.
>>I see a tautology (in the
>>sense meant by the original poster) as a process that you can turn
>>around, and it would still be the same. You can do that with a=b
>>-> b=a, but you can't do it with
>>...-->F1-->(survival)-->(reproduction)-->F2-->...
>
>
>Oh, but you can. The key here is how we _measure_ fitness. Start with
>the statement "survival of the fittest". This basically states that
>survival = fittest
>But how do we measure fitness? With survival of course! Then
>we have that
>fitness = survival
I'd have to take issue with this bit, and say that these are not
reversible equalities. Survival is based on certain qualities
which may be collectively grouped as 'fitness', but 'fitness' is
something more than just the fact that the organism has survived.
What do I mean? I mean that you can take a group of organisms and
watch them and see the qualities of the survivors, and determine
that as fitness. However, you cannot take the survivors, and from
them work out the characteristics of the initial cohort. You can't
work out what the differences were that made them survive while
the others fail.
In other words, it's one-way. I really think that this tautology
business is misplaced.
>Hence, a tautology in the sense of the original poster. As long as we
>don't treat fitness any different from survival then the statement is
>a tautology.
Mmmm, but as I hope I've demonstrated, fitness and survival are
*not* equalities... they're consecutive elements in the iteration.
Onar Aam
>I'd have to take issue with this bit, and say that these are not
>reversible equalities. Survival is based on certain qualities
>which may be collectively grouped as 'fitness', but 'fitness' is
>something more than just the fact that the organism has survived.
In that case you are using fitness in a different way than is common
among evolutionary biologists. In computer simulations and
experiments we frequently see fitness defined as a scalar. (e.g.
fitness = 0.92) But very rarely is that value specified as anything
more than reproductive success. When used in this way, fitness is
completely indistinguishable from the concept of survival.
Onar.
Thomas K. Dibenedetto
: Onar Aam (on...@hsr.no) wrote:
: : Oh, but you can. The key here is how we _measure_ fitness. Start with
: : the statement "survival of the fittest". This basically states that
: : survival = fittest
: fittest ) survival
: where ) means "if .. then". If fittest, then survival.
etc. etc. snip snip
Sorry if I've missed something guys, but what is this all about? Reading
the thread title, I sorta kinda expected a discussion of Darwins theory,
not a discussion of a catchy little synopsis that Spencer used to
popularize D's theory for the unthinking masses. Whether SotF is a
tautology or not doesnt answer the question posed in the title, and is,
from a scientific perspective, an uninteresting question.
It is an abstraction of D's theory, used for explanatory purposes.
Clearly it causes more confusion than illumination, and should be
abandonded (which it has been for the most part...I cant remember ever
even hearing the phrase in an academic setting, except in derision).
--
Tom diBenedetto
Fish Division
University of Michigan Museum of Zoology
Shane McKee
>>I'd have to take issue with this bit, and say that these are not
>>reversible equalities. Survival is based on certain qualities
>>which may be collectively grouped as 'fitness', but 'fitness' is
>>something more than just the fact that the organism has survived.
Onar Aam wrote:
>In that case you are using fitness in a different way than is common
>among evolutionary biologists. In computer simulations and
>experiments we frequently see fitness defined as a scalar. (e.g.
>fitness = 0.92) But very rarely is that value specified as anything
>more than reproductive success. When used in this way, fitness is
>completely indistinguishable from the concept of survival.
Simply because fitness may (in some models) be conveniently defined
*in terms of* survival, this does not make the two concepts the
same thing, as you seem to be arguing above, and certainly, as I
have elaborated before, the process Fitness-->survival is not
reversible.
You suggest that fitness is rarely defined as anything more than
reproductive success, but this is a misleading simplification. The
fitness of an allele can, to be sure, be given a numerical value
denoting the proposed contribution it makes to its own survival or
propagation through the gene pool, but simply because an allele
survives, you cannot automatically deduce fitness.
As I've said, it's a one-way process.
What I didn't mention, but which is perhaps important, is that
while fitness can be given a value, its conversion to survival is
based on probability and not absolutes.
Shane
__
/ \
/ 0 ] Shane McKee, Northern Ireland
\/\_/
Chris Colby
thing. Arguing whether the phrase "survival of the fittest" is a
tautology is stupid if the intention is to criticise evolutionary
biology. This is just a bumper sticker phrase, not a part of the
theory of evolution. You might as well try to criticise Einstein's
theory of relativity by tearing apart the phrase "everything's
relative".
Evolution is defined as "a change in allele frequencies". (Alleles
are alternate forms of the same gene, like the alleles for brown
eyes versus blue eyes). Natural selection is one of the mechanisms
of evolution; it is defined as differential reproductive success
of genotypes. Genotypes that have a higher reproductive success
(i.e. leave more offspring) increase in frequency in the population.
Fitness is simply the measure of reproductive success.
Fitness has many components, of which survival is only one. Fertility,
fecundity and mating success are other important components of fitness.
Whether or not the phrase "survival of the fittest" is a tautology,
fitness is not equivalent to survival in mathematical models of
evolutions.
Taking the time to understand the basics of evolution is a rewarding
process. If you have the time to post your opinions, you should have
time to learn something about evolution. Knowledge of evolution will
not arrive at your brain via osmosis; it is not something that can be
figured out by pondering phrases like "survival of the fittest" or
baselessly speculating about the "adaptive significance" of some
trait. You have to spend some time reading the works of people who
study it. Not only will your own understanding of evolution increase
(and this should be reward enough), sci.bio.evolution would benefit
if the "evolutionary discussions" progressed beyond the sophomoric
levels evidenced by this thread.
Chris Colby --- email: co...@biology.bu.edu or co...@acs.bu.edu ---
"'My boy,' he said, 'you are descended from a long line of determined,
resourceful, microscopic tadpoles--champions every one.'"
--Kurt Vonnegut from "Galapagos"
Onar Aam
Chris Colby writes:
>This whole thread goes to show that a little knowledge is a dangerous
>thing. Arguing whether the phrase "survival of the fittest" is a
>tautology is stupid if the intention is to criticise evolutionary
>biology.
As I've stated before, I'm not criticising evolutionary biology. All I
have done is to point out that evolutionary biology does in fact have
a tautological element to it ... as do all major scientific theories.
And as I've said before: there is nothing wrong about that.
The crux here is (as Shane McKee pointed out) that survival is a one-time
event while fitness is an overall, probabilistic value. Hence you cannot
deduce from a single incident of reproducitive success that some organism
is fit. It may have been extremely lucky, a momentary fortunate
environment etc. However, after X (say 10,000) generations you can say
something about fitness because the odds for a lucky strike decreases for
every new generation. In other words, reproduction is a fuzzy function
and only by iterating the function over time will fitness be revealed.
This means that fitness is not merely survival, it is the property of
being _good_ at survival.
Shane McKee pointed out that we cannot deduce fitness from single
incidents of survival, but the reverse is also true. You cannot deduce
survival from fitness. A fit organism may be extremely unlucky and
for some reason die before it manages to reproduce. (it may be struck
by lightening, hit by a meteor etc.) In other words, due to the fuzzy
nature of reproduction it makes no sense to compare survival and
fitness in individual iterations.
But my point in all this is that survival is a fuzzy function of fitness
and at the same time that our measure of fitness is a function of
survival. This is an unavoidable circularity which arises from the
very nature of the phenomenon we are studying.
Onar.
Alex Merz
[...]
>But my point in all this is that survival is a fuzzy function of fitness
>and at the same time that our measure of fitness is a function of
>survival. This is an unavoidable circularity which arises from the
>very nature of the phenomenon we are studying.
Tautology implies circular _closure_. The T of E is _not_ closed
in a logical sense - it has inputs and outputs.
The theory _is_ iterative and recursive; but it is not tautological.
-Alexey
>Onar.
Onar Aam
>Tautology implies circular _closure_. The T of E is _not_ closed
>in a logical sense - it has inputs and outputs.
>
>The theory _is_ iterative and recursive; but it is not tautological.
The concept of tautology naturally extends into fuzzy logic (which of
course gives us _degrees_ of tautology). Since fitness in its nature
is a fuzzy concept the statement "survival of the fittest" cannot be
a pure tautology. However, taking this into account we find that the
statement to a large *degree* is tautological. It is in this fuzzy
sense that most (if not all) scientific theories are tautological.
Onar.
Benjamin H. Diebold
the idea of degrees of tautology. Looks like lost-cause wiggling to me.
Anyway, Stephen J. Gould has a nice article on exactly this topic. It
first appeared in Harper's Magazine in 1976, in response to an article by
a creationist who claimed Darwinism was dead because it's a tautology.
The response was expanded by Gould into an article, and published as
chapter 8 of Laurie Godfrey's _Scientists confront creationism_ (1983).
Gould neatly dispenses with this "Darwinism as tautology" idea,
essentially because "criteria of fitness independent of survival can be
applied to nature, and have been used consistently by evolutionists" (p.
143). I recommend the article (and the book) to interested readers who
want details.
Ben
Karl
: >Tautology implies circular _closure_. The T of E is _not_ closed
It seems to me that the question of tautology hinges on which
part of Darwinism you look at. If you look at the definition
of "fitness" (or the definition of any term, for that matter),
you will find that the definition is a tautology.
This should not be too surprising, since -=ALL=- definitions
are tautologies. Logicall, A = A is always true.
Darwinian evolutionary theories, on the other hand, are not
tautological. They make statements along the lines of"
"If this theory is true, then we should observe the following".
Taking the example of fitness, which is the favorite point of
attack in the "tautology" argument, we have:
1) A variable called "fitness", which is defined in a specific
manner. (Specifically, it's defined as the number of offspring
which live to reproduce.)
2) A statement that given two groups of organisms with different
average fitnesses, over the course of generations, the group with
a higher value of "fitness" will be more numerous than the group
with the lower value.
This is not a tautology because it can easily work the other way.
Upon examination after several generations, it could turn out that
the group with a lower fitness would have displaced the group with
the higher fitness. In this case, "survival of the fittest" would
have to be discarded.
............Karl
Alex Merz
>>in a logical sense - it has inputs and outputs.
>>
>>The theory _is_ iterative and recursive; but it is not tautological.
>
>
>course gives us _degrees_ of tautology). Since fitness in its nature
>is a fuzzy concept the statement "survival of the fittest" cannot be
>a pure tautology. However, taking this into account we find that the
>statement to a large *degree* is tautological. It is in this fuzzy
>sense that most (if not all) scientific theories are tautological.
This is incorrect and misleading.
Logically, a statement is tautological or it is not. In evolutionary
theory fitness is _un_ambiguously defined as relative reproductive
success. This is not fuzzy. The theory is not tautological -
not even a little!
-Alexey
Grant B. Harris
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 1.22 (Windows; I; 16bit)
Status: R
Even if we assume that Darwinian theory is not tautological, isn't there
still the problem of the origin of the first replicating entity with
hereditary variation? Some would argue that Darwinism "assumes what it
seeks to explain". Without replicating entities, there can be no
'natural selection', so how can a relicating entity evolve in the first
place? Thoughts?
-- Grant Harris
Peter Nyikos
Status: R
me...@ohsu.EDU (Alex Merz) writes:
>In article <4b48r4$9...@nntp5.u.washington.edu>, Onar Aam <on...@hsr.no> wrote:
>>
>>>in a logical sense - it has inputs and outputs.
>>>
>>>The theory _is_ iterative and recursive; but it is not tautological.
>>
>>
>>course gives us _degrees_ of tautology). Since fitness in its nature
>>is a fuzzy concept the statement "survival of the fittest" cannot be
>>a pure tautology. However, taking this into account we find that the
>>statement to a large *degree* is tautological. It is in this fuzzy
>>sense that most (if not all) scientific theories are tautological.
Although I am a mathematician well versed in logic, I have no
problem with such loose informal uses of "tautological" as long
as there is no misunderstanding of what is meant.
>This is incorrect and misleading.
>Logically, a statement is tautological or it is not. In evolutionary
>theory fitness is _un_ambiguously defined as relative reproductive
>success. This is not fuzzy. The theory is not tautological -
>not even a little!
Unambiguously? Just how is relative reproductive success defined?
Do you go exactly one generation from where you are? Do you
count the number of offspring who survive to reproductive age,
or just the ones that reproduce--and what do you mean by THEM
reproducing? Do they have to have no more than one offspring
of their own surviving to reproductive age?
Oh, and by the way, what do you mean by "survival" in this
context? Do you mean survival of strains for one more generation?
If so, "survival of the fittest" is pretty darn close to being
a tautology in the literal sense!
Peter Nyikos -- standard disclaimer --
Professor, Dept. of Mathematics
University of South Carolina
Columbia, SC 29208
Matthew P Wiener
NNTP-Posting-Host: sagi.wistar.upenn.edu
In-reply-to: me...@ohsu.EDU (Alex Merz)
Status: R
In article <4bo7sp$3...@fremont.ohsu.edu>, merza@ohsu (Alex Merz) writes:
>Logically, a statement is tautological or it is not. In evolutionary
>theory fitness is _un_ambiguously defined as relative reproductive
>success. This is not fuzzy. The theory is not tautological - not even
>a little!
It is a priori *impossible* for "survival of the fittest" to be a
tautology, no matter how compelling one makes the argument.
It is logically possible that, for unconsidered reasons having nothing
to do with "tooth and claw", the not-so-fit will also survive. Before
Darwin, it was routinely believed that there were no extinctions since
the Biblical flood. The known examples like the dodo could be written
off as "not a kind" or a belief that there were dodos elsewhere. The
nonmodern fossil species were automatically given antediluvian status.
Presumably divine intervention always saved the last few or something
to that effect. The truth of "survival of the fittest" thus gives us
positive information regarding the reductionist assumptions behind the
usual arguments.
In physics, one might think that having *defined* velocity as the
derivative of distance with respect to time, that it is a *tautology*
that distance is just the integral of velocity over time. After all,
is this not just pure mathematical deduction, Calculus 101?
The ultimate answer is no. The real world doesn't have to accord
with our simple Cartesian model for basic physics. That it does is
an experimental issue, not a logical one.
--
-Matthew P Wiener (wee...@sagi.wistar.upenn.edu)
Peter Nyikos
>NNTP-Posting-Host: sagi.wistar.upenn.edu
>In-reply-to: me...@ohsu.EDU (Alex Merz)
>Status: R
>In article <4bo7sp$3...@fremont.ohsu.edu>, merza@ohsu (Alex Merz) writes:
>>Logically, a statement is tautological or it is not. In evolutionary
>>theory fitness is _un_ambiguously defined as relative reproductive
>>success. This is not fuzzy. The theory is not tautological - not even
>>a little!
>It is a priori *impossible* for "survival of the fittest" to be a
>tautology, no matter how compelling one makes the argument.
I'd say that depends strongly on how one defines "survival" and
"fittest", as I pointed out in my follow-up to Alexey Merz.
>It is logically possible that, for unconsidered reasons having nothing
>to do with "tooth and claw", the not-so-fit will also survive.
That depends on where you draw the line between the fit and
not-so-fit. Also on where you draw the lines between "fit",
"fitter", and "fittest".
Before
>Darwin, it was routinely believed that there were no extinctions since
>the Biblical flood. The known examples like the dodo could be written
>off as "not a kind" or a belief that there were dodos elsewhere.
This is the first I've heard of such a thing. Can you steer
me to a reference?
>Presumably divine intervention always saved the last few or something
>to that effect. The truth of "survival of the fittest" thus gives us
>positive information regarding the reductionist assumptions behind the
>usual arguments.
I'm having trouble following this. Can we back up a bit? From
what Alexey Merz said, we might define terms as follows:
fit = having a relative reproductive advantage
survival = having descendents over a significant [and we
could haggle over how many is "significant"] number of generations
fittest = those whose relative reproductive advantage passes
a certain given point.
With the above definitions, the truth of "survival of the
fittest" depends crucially on just what that "certain given
point" is. And it can be a really barren tautology if you
put that certain point too high, by excluding the vast majority
of fit creatures that also survive.
Matthew P Wiener
Darwin's own theory, as per the subject line. But SOTF is the item
of discussino.
------------------------------------------------------------------------
In article <4ch4un$2...@nntp5.u.washington.edu>, nyikos@math (Peter Nyikos) writes:
>Reply-To: wee...@sagi.wistar.upenn.edu (Matthew P Wiener)
>>>Logically, a statement is tautological or it is not. In evolutionary
>>>theory fitness is _un_ambiguously defined as relative reproductive
>>>success. This is not fuzzy. The theory is not tautological - not even
>>>a little!
>>It is a priori *impossible* for "survival of the fittest" to be a
>>tautology, no matter how compelling one makes the argument.
>I'd say that depends strongly on how one defines "survival" and
>"fittest", as I pointed out in my follow-up to Alexey Merz.
As long as you define them with reference to the real world, then it
is not a tautology.
>>It is logically possible that, for unconsidered reasons having nothing
>>to do with "tooth and claw", the not-so-fit will also survive.
>That depends on where you draw the line between the fit and
>not-so-fit. Also on where you draw the lines between "fit",
>"fitter", and "fittest".
First off, I do not interpret SOTF as a logical proposition, but an
attempt to say something non-trivial about the real world. (In particular,
no one who claims it is true believes it is a specialization of survival
of everything.)
So regardless of where the lines are drawn, SOTF, in addition to being
some claim about who survives and who does not, also claims that there
are species that went extinct.
*That* is an experimental observation.
>>Before Darwin, it was routinely believed that there were no
>>extinctions since the Biblical flood. The known examples like the
>>dodo could be written off as "not a kind" or a belief that there
>>were dodos elsewhere.
>This is the first I've heard of such a thing. Can you steer me to a
>reference?
And God spoke to Noah, and to his sons with him, saying,
And behold, I establish my covenant with you, and with your
seed after you,
And with every living creature that is with you, of the birds,
of the cattle, and of every beast of the earth with you, from
all that came out of the ark, to every beast of the earth.
And I will establish my covenant with you, neither shall all
flesh be cut off any more by the waters of the flood, neither
shall there any more be a flood to destroy the earth.
And God said, this is the token of the covenant which I make
between me and you and every living creature that is with
you, for perpetual generations: [Genesis 9:8-12]
Some people take that "perpetual" pretty seriously. See, for example,
Ellen G White PATRIARCHS AND PROPHETS.
>>Presumably divine intervention always saved the last few or something
>>to that effect. The truth of "survival of the fittest" thus gives us
>>positive information regarding the reductionist assumptions behind the
>>usual arguments.
>I'm having trouble following this.
What I mean is that the default background assumption throughout science
seems to be nonstop reductionism. While a roaring success story, it
remains an empirical, not a logical question.
Did you understand my example regarding the integral of velocity gives
you distance is _not_ a tautology, in spite of this being a trivial
calculus 101 result?
It's much easier to discuss that example, since there are universally
agreed upon definitions, the mathematical proof is universally agreed
upon, and yet I still claim the final result is not a tautology. And
if non-tautology holds for a _precise_ example, then for sure it also
holds for a fuzzier example.
> Can we back up a bit? From
>what Alexey Merz said, we might define terms as follows:
>[omitted]
>With the above definitions, the truth of "survival of the fittest"
>depends crucially on just what that "certain given point" is.
Its truth may depend on such, maybe, but not its alleged tautological
nature, which is what I am commenting on. If observation reveals that
species do not go extinct, then SOTF--in the non-trivial sense that I
mentioned above--is an experimentally falsified statement, *regardless*
of which particular definition you give or where you draw the lines
between fit, fitter, fittest, etc.
Since it is conceivably falsifiable, it is not tautological.
> And it can be a really barren tautology if you
>put that certain point too high, by excluding the vast majority
>of fit creatures that also survive.
I view SOTF as something that should have a stochastic interpretation,
and that the question of where the lines are is void.
Onar Aam
Status: R
>>course gives us _degrees_ of tautology). Since fitness in its nature
>>is a fuzzy concept the statement "survival of the fittest" cannot be
>>a pure tautology. However, taking this into account we find that the
>>statement to a large *degree* is tautological. It is in this fuzzy
>>sense that most (if not all) scientific theories are tautological.
>
>This is incorrect and misleading.
>
>Logically, a statement is tautological or it is not.
Ever heard of Fuzzy Logic? It is as valid as any logic, and when we are
dealing with probabilistic (fuzzy) concepts like fitness it is sensible
to use fuzzy logic. If this is hard to comprehend then consider this:
Fitness does not say if an organism will reproduce or not. A fit
organism may be struck by lightening and thereby fail to reproduce while
an unfit organism may miraculously manage to reproduce. In other words,
fitness does not say anything about survival, but rather of the
_likeliness_ of survival. A well adapted organism has a greater odds
of reproductive success than does a poorly adapted organism. Hence, fitness
is not an absolute concept, but a probabilistic or FUZZY concept. Stated
in words fitness may therefore be defined as a "tendency to survive". Then
survival of the fittest translates into the following: survival of those
that tend to survive. Now, had it not been for the word "tend" this would
have been a good ol' tautology, a non-informative statement. But precisely
because "tend" is a fuzzy concept the statement becomes a fuzzy tautology.
It's as simple as that.
Now, the clue about the SOTF is not what it says but what it _doesn't_ say,
namely that those that are unfit will tend NOT to survive. In other words,
SOTF describes a _differentiation_ process, a natural filter. This is the
point where things start to become highly non-obvious, because the natural
question to ask next is: what passes through this natural filter? And it is
basically *this* question evolutionary biologists are concerned with. The
reason this is highly non-obvious is because it is not limited to biology,
but concerns everything in the universe which tend to stick around (e.g.
physical particles, consciousness). A thorough answer to this question will
therefore require a complete theory of perpetual motion. In biology there
is only one branch which seriously tries to deal with this phenomenon,
namely the autopoiesis camp.
Onar.
Onar Aam
>>>It is a priori *impossible* for "survival of the fittest" to be a
>>>tautology, no matter how compelling one makes the argument.
>
>>I'd say that depends strongly on how one defines "survival" and
>>"fittest", as I pointed out in my follow-up to Alexey Merz.
>
>As long as you define them with reference to the real world, then it
>is not a tautology.
But then you're shutting out a whole field of inquiry. It's like a theory
of all colors except green. Logic and mathematics are not somehow
mysteriously separate from the real world.
>The ultimate answer is no. The real world doesn't have to accord
>with our simple Cartesian model for basic physics.
Oh, but it does. There are literally infinitely many ways the universe
could have been, but for some strange reason it has a structure which
happens to be very simple. The universe could have been governed by
2,976th order differential equations, but it is not. It is governed by
2nd order differential equations. Why? We don't know, but we do know that
our simple Cartesian models are extremely efficient at describing the
universe, and this requires an explanation. (Known as Wigner's problem)
Onar.
Peter Nyikos
>As a side comment, let me note that "survival of the fittest" is not
>Darwin's own theory, as per the subject line. But SOTF is the item
>of discussino.
>In article <4ch4un$2...@nntp5.u.washington.edu>, nyikos@math (Peter Nyikos) writes:
>>Reply-To: wee...@sagi.wistar.upenn.edu (Matthew P Wiener)
>>>In article <4bo7sp$3...@fremont.ohsu.edu>, merza@ohsu (Alex Merz) writes:
>>>>Logically, a statement is tautological or it is not. In evolutionary
>>>>theory fitness is _un_ambiguously defined as relative reproductive
>>>>success.
I pointed out some ambiguities and have yet to see Alex or
anyone else address them.
This is not fuzzy. The theory is not tautological - not even
>>>>a little!
This is a *non sequitur* since Alexey never gave the definition
of "survival". Someone could easily say, by giving the "right"
definitions to these things,
Species A was more fit at a given real point of time, t,
than Species B, because Species A went on for at least
one more generation, and Species B did not. IN OTHER
WORDS, Species A survived and Species B did not.
Thus they could be expressing their belief that "survival of
the fittest" is indeed a tautology.
>>>It is a priori *impossible* for "survival of the fittest" to be a
>>>tautology, no matter how compelling one makes the argument.
>>I'd say that depends strongly on how one defines "survival" and
>>"fittest", as I pointed out in my follow-up to Alexey Merz.
>As long as you define them with reference to the real world, then it
>is not a tautology.
I disagree, see above. For "Species A" one can (for instance)
put *Homo sapiens*, for
Species B put the passenger pigeon, and t can be a certain
point around the previous turn of the century which I can look
up if anyone is interested. Each relevant to which was "fit"
at time t, and FOR THE SAME REASON, which survived, if one
uses the "right" definitions for making SOTF a tautology.
>>>It is logically possible that, for unconsidered reasons having nothing
>>>to do with "tooth and claw", the not-so-fit will also survive.
Note that the highly reductionist definition of "fitness" up
there makes NO mention of "tooth and claw". Is this definition
of fitness the one you wish to adopt too? The biology books I've
seen which define fitness all seem to endorse it.
Once you adopt it, one can simply define the statement
"the not-so-fit will also survive" into self-contradiction:
>>That depends on where you draw the line between the fit and
>>not-so-fit. Also on where you draw the lines between "fit",
>>"fitter", and "fittest".
>First off, I do not interpret SOTF as a logical proposition, but an
>attempt to say something non-trivial about the real world.
Laymen often interpret SOTF a lot more broadly than the definition
"relative reproductive success", as suggested by your own use
of "tooth and claw". Do you wish to leave the door open to
such broadenings of the term?
>So regardless of where the lines are drawn, SOTF, in addition to being
>some claim about who survives and who does not, also claims that there
>are species that went extinct.
I beg to differ. The claim that there are species that went
extinct is an extra datum which, if one wishes to define terms
so that SOTF is a tautology, is simply empirical fact that
ADDS to one side of the tautology, and hence to both sides.
[Here I have deleted something which I dealt with in a separate
post; if the moderators rule it off-topic, I will post it to
talk.origins.]
>>>Presumably divine intervention always saved the last few or something
>>>to that effect. The truth of "survival of the fittest" thus gives us
>>>positive information regarding the reductionist assumptions behind the
>>>usual arguments.
>>I'm having trouble following this.
>What I mean is that the default background assumption throughout science
>seems to be nonstop reductionism. While a roaring success story, it
>remains an empirical, not a logical question.
Not if the reductionism proceeds nonstop, as in my indented
statement comparing Species A with Species B.
>Did you understand my example regarding the integral of velocity gives
>you distance is _not_ a tautology, in spite of this being a trivial
>calculus 101 result?
Sure, but this is sci.bio.evolution, not talk.philosophy.misc
nor sci.philosophy.tech,
which is where this line of talk inevitably leads us: Imannuel
Kant, "analytic vs. synthetic", "synthetic a priori" vs "tautological",
etc. At bottom, this last distinction is a feature of psychology,
not of logic, IMO, and I'm told Quine feels this way about it too.
Be that as it may, in the case of SOTF everything depends on
how the terms are defined, IMNSHO.
>> Can we back up a bit? From
>>what Alexey Merz said, we might define terms as follows:
>>[omitted]
>>With the above definitions, the truth of "survival of the fittest"
>>depends crucially on just what that "certain given point" is.
>Its truth may depend on such, maybe, but not its alleged tautological
>nature, which is what I am commenting on. If observation reveals that
>species do not go extinct, then SOTF--in the non-trivial sense that I
>mentioned above--is an experimentally falsified statement,
I'd have to see what that non-trivial sense is before I agree to
this.
>I view SOTF as something that should have a stochastic interpretation,
>and that the question of where the lines are is void.
Please elucidate that stochastic interpretation.
Matthew P Wiener
>>>I'd say that depends strongly on how one defines "survival" and
>>>"fittest", as I pointed out in my follow-up to Alexey Merz.
>>As long as you define them with reference to the real world, then it
>>is not a tautology.
>But then you're shutting out a whole field of inquiry. [...]
So it goes in the world.
>>The ultimate answer is no. The real world doesn't have to accord
>>with our simple Cartesian model for basic physics.
>Oh, but it does. There are literally infinitely many ways the universe
>could have been, but for some strange reason it has a structure which
>happens to be very simple. [...]
Which means, like I said, the ultimate answer is "not a tautology".
Matthew P Wiener
n>wee...@sagi.wistar.upenn.edu (Matthew P Wiener) writes:
>>In article <4ch4un$2...@nntp5.u.washington.edu>, nyikos@math (Peter Nyikos) writes:
>>>Reply-To: wee...@sagi.wistar.upenn.edu (Matthew P Wiener)
>>>>It is a priori *impossible* for "survival of the fittest" to be a
>>>>tautology, no matter how compelling one makes the argument.
>>>I'd say that depends strongly on how one defines "survival" and
>>>"fittest", as I pointed out in my follow-up to Alexey Merz.
>>As long as you define them with reference to the real world, then it
>>is not a tautology.
> Species A was more fit at a given real point of time, t,
> than Species B, because Species A went on for at least
> one more generation, and Species B did not. IN OTHER
> WORDS, Species A survived and Species B did not.
>I disagree, see above. For "Species A" one can (for instance) put
>*Homo sapiens*, for Species B put the passenger pigeon, and t can be
>a certain point around the previous turn of the century which I can
>look up if anyone is interested. Each relevant to which was "fit" at
>time t, and FOR THE SAME REASON, which survived, if one uses the
>"right" definitions for making SOTF a tautology.
_Obviously_ if you give "fit" and "survive" the same definition, you
have a tautology. Try giving semirealistic definitions.
>>>>It is logically possible that, for unconsidered reasons having nothing
>>>>to do with "tooth and claw", the not-so-fit will also survive.
>Note that the highly reductionist definition of "fitness" up
>there makes NO mention of "tooth and claw". Is this definition
>of fitness the one you wish to adopt too? The biology books I've
>seen which define fitness all seem to endorse it.
Try defining fitness locally--per creature, one generation--and survival
via the long term--per species, say over thousands of generations.
As is, I've never seen fitness defined as merely surviving to reproduce.
>Once you adopt it, one can simply define the statement
>"the not-so-fit will also survive" into self-contradiction:
Huh? No such self-contradiction is derived.
>>>That depends on where you draw the line between the fit and
>>>not-so-fit. Also on where you draw the lines between "fit",
>>>"fitter", and "fittest".
>>First off, I do not interpret SOTF as a logical proposition, but an
>>attempt to say something non-trivial about the real world.
>Laymen often interpret SOTF a lot more broadly than the definition
>"relative reproductive success", as suggested by your own use
>of "tooth and claw". [...]
Eh? My own use was only suggestive.
>>So regardless of where the lines are drawn, SOTF, in addition to being
>>some claim about who survives and who does not, also claims that there
>>are species that went extinct.
>I beg to differ. The claim that there are species that went extinct
>is an extra datum which,
As I said, if no species go extinct, SOTF is a pointless utterance to
even consider.
> if one wishes to define terms so that SOTF
>is a tautology,
If you wish to do that, I have zero interest in discussing anything.
Coming up with the trivial is for the trivial.
> is simply empirical fact that ADDS to one side of the
>tautology, and hence to both sides.
Eh? How does it add to either side?
>>>>Presumably divine intervention always saved the last few or something
>>>>to that effect. The truth of "survival of the fittest" thus gives us
>>>>positive information regarding the reductionist assumptions behind the
>>>>usual arguments.
>>>I'm having trouble following this.
>>What I mean is that the default background assumption throughout science
>>seems to be nonstop reductionism. While a roaring success story, it
>>remains an empirical, not a logical question.
>Not if the reductionism proceeds nonstop, as in my indented
>statement comparing Species A with Species B.
Your statement simply considered "fitness" as a synonym for "survival".
It had nothing to do with reductionism.
>>Did you understand my example regarding the integral of velocity gives
>>you distance is _not_ a tautology, in spite of this being a trivial
>>calculus 101 result?
>Sure, but this is sci.bio.evolution, not talk.philosophy.misc nor
>sci.philosophy.tech, which is where this line of talk inevitably
>leads us: Imannuel Kant, "analytic vs. synthetic", "synthetic a
>priori" vs "tautological", etc. [...]
Leads to???? That's where we _are_! So try answering my question.
>>Its truth may depend on such, maybe, but not its alleged tautological
>>nature, which is what I am commenting on. If observation reveals that
>>species do not go extinct, then SOTF--in the non-trivial sense that I
>>mentioned above--is an experimentally falsified statement,
>I'd have to see what that non-trivial sense is before I agree to this.
What I stated.
>>I view SOTF as something that should have a stochastic interpretation,
>>and that the question of where the lines are is void.
>Please elucidate that stochastic interpretation.
Lots of models exist. For example, give creatures probability distributions
(that are a themselves a probabilitistic function of the environment) that
predict how many offspring they will have per time period, plus model the
environment itself, partly as a background, and partly dependent on the
creatures themselves. Throw in some energetic constraints and the like,
mortality considerations, and so on.
In such models, there is no absolute notion of fitness, even when the numbers
have definite meaning according to the program. So I don't expect the real
word to have clear notions of the "fittest".
Onar Aam
Peter Nyikos wrote:
>I pointed out some ambiguities and have yet to see Alex or
>anyone else address them.
I want to applaud Peter Nyikos for his crisp understanding of the
essence of this thread. I will try to elaborate on the ambiguities
with the concept of fitness which he has correctly pointed out.
The general confusion and ambiguities concerning fitness is due to
a whole army of hidden assumptions in biology. Scientists can afford
to neglect these conditions because they just "work". I will try to
summarize their most important aspects in two statements.
--- An organism is fit by default
Consider the statement "X is fit". As we see fitness is a claim
about X, a property that X posesses. Obviously X cannot be
everything that springs to mind. Neither a bottle or a telescope
can be fit. Therefore a natural question to ask is what values the
variable X can have. A quick and simple answer is the Organism.
But what the heck is an organism!? The answer to this question is
everything but trivial. In fact, a thorough answer will be a _theory_
of the organism. Basically an organism is a survival machine. It's
intrinsic organization is that of _perpetuation_. So of all possible
physical systems the organism belongs to the small class of systems
that embodies perpetual motion (*). But note that this from the very
start gives the biologist a lot of fitness for free. An organism is
by definition a machine that tends to survive, and therefore has a
high degree of fitness by default, before any formal concept of fitness
has been defined. So by the very act of limiting X to organisms we have
already defined fitness in a very broad sense. In short, the technical
concept of fitness assumes that it is dealing with a survival machine
of some sort. This initial fitness leverage that a survival machine
yields is one very important hidden assumption in the concept of
fitness.
But science rarely leaves any room for broad concepts. In order
to be the subject of mathematical models the organism needs to be
precisely defined which is done by operationally defining the organism
as a _replicator_. The technical definition of fitness is therefore a
function of replication.
--- Fitness is a function of measurement
In a previous post I showed that natural selection may be seen as a
discrete filter process. This implies that natural selection can
be described systemically in terms of filter theory and discrete
signal processing. If you are familiar with these fields then you
will immediately know that there are a myriad of factors that can
influence the measurement of such a system. The concept which is most
relevant to evolution is something which is called the "measurement
window" (this concept has many names but the term "window" is very
common). This window is the range in which the measurement takes place
As I will demonstrate, the size of that window is a parameter that will
greatly affect the measured outcome. Let me give you an example from
traffics:
Imagine a car queue. We define the Observation Window as the range that
each driver can observe in front of him. Suppose now that the car is
driving at a constant speed, say 40 m/h. What would happen if one of the
cars suddenly reduced his speed to 30 m/h? That speed reduction would
spread backwards in the queue like a schock wave. Let us now treat that
wave as a harmonic signal. This will have several properties such as
amplitude and frequency. The frequency corresponds to the abruptness
of change in the car queue. The more abrupt the change is, the higher
frequency the signal has. Now, what affect does the width of the
observation window have on this signal? Suppose each driver can only
see one car in front of him. Then he will not know that there is a
wave of speed reduction approaching him before the car in front of him
starts slowing down. So not before he sees the break lights of the car
in front of him will he himself step on the breaks. The wavelength of
this speed reduction wave will therefore be very small (and hence the
frequency very high). But suppose we increased the width of the
observation window to, say, 10 cars instead of one. Then each driver can
respond to the wave much earlier and thereby make the change less abrupt.
Thus, increasing the width of the observation window tends to filter away
the high frequencies, effectively acting as a low-pass filter. This
illustrates how measurement can affect the signal.
End of detour. Back to evolution. Suppose now that we did the same thought
experiment for the measurement of fitness. Let us define a similar
window parameter, W, as the number of generations over which we measure
fitness. Let us now conceptually "tune" W from its minimal value (one
generation) up to its maximum value (infinitely many generations) and see
how it affects fitness. Starting with W=1 we will see a lot of
fluctuations. That is, the resulting signal will be fuzzy. If we increase
W to 10 we will filter out most of the high frequent noise and we will see
the more low-frequent change which evolutionists refer to as "adaption" or
micro-evolution. If we increase W to 10,000 then we will weed out much of
this micro-evolution. For instance, the famous example of the increase in
the black pepper moth population versus the white moth population would not
turn up in the recording process at W=10,000. From this we can conclude that
much of what is called evolution today is in fact species _homeostasis_. For
low values of W we see the adaption, at high values of W we see the stasis.
At even higher values of W we start recording extinction and speciation
dynamics etc. etc.
Alex Merz wrote that "in evolutionary theory fitness is
_un_ambiguously defined as relative reproductive success." I hope I have
demonstrated that this is a naive conception of fitness. Fitness is a
compound concept, governed by numerous parameters. If I were to split
fitness into the above discussed components I would do it somewhat like
this:
Fitness =
60% initial fitness leverage of the survival machine
30% measurment parameter distortion
10% other factors
But I hasten to add that this is only a rough estimate and that the
weighting of the components depends on the values of the parameters.
Furthermore, this estimate probably suffers from a great flaw. As
behavioral psychologists would say: behavior is 100% genetic and
100% environmental. In other words, it may not be possible to
meaningfully split up a concept into distinct components because
those components will always be interdependent and overlapping. But
this only strengthens the notion that fitness is an ambiguous concept.
(*) By perpetual motion I do not mean perpetual motion machines of the 1st
and 2nd order, but a general notion of systems that tend to self-perpetuate.
Onar.
Thomas K. Dibenedetto
[much ado about fitness]
I fail to see the relevance of long explanations of easily made points,
namely that "fitness" can be ascribed to the players in evolutionary
history at any number of scales. I dont find this relevant to a
discussion of Darwinism, because "fitness" in such a sense is focused on
one particular scalar level; namely the relative propensity for survival
(and propogation) of particular variants in a population. It is forever
possible to seize upon all of the potential meanings and connotations of
a word, explicate them, and thereby demonstrate that any particular use
of the word can easily be exploded into meaninglessness. It seems to me an
empty exercise; what is relevant is how the word is actually being used
in the specific example under discussion.
Clearly appeals to reason (postings in this thread which have
pointed out the SotF is a popular summary of D's theory, rather than its
formal explication) have fallen on deaf ears. If there is a fascination
with analyzing this aphorism, so be it. Can we all agree that there are
variants within populations? And that over the course of many
generations, some of these variants persist, whereas others disappear?
This is what "survival" refers to; not the survival of an individual, but
the persistence of a variation. This differential persistance is posited
to be non-random; to be correlated with a certain efficiency in
utilizing the resources of the environment in which the variants find
themselves. This efficiency is seen (in a theoretical sense) as a measure
of adaptation. It is the idea behind the concept "fitness".
Attempts to measure this fitness are part of the grand research
project of evolutionary biology. We can look at the raw persistance of the
variants and _identify_ the distribution of relative fitness, because we are
operating within the assumptions of the theory (the "survivors" are
_expected_ to be the fit ones in the theory). Thus we tend to measure
fitness by looking at reproductive success, for clearly this is the
engine behind persistance. Such identifications do not test the validity
of the theory; they are not intended to. This exercise is part of
_doing_ evo biology, not testing its assumptions. It is only those who
wish to question the theory (mainly creationists) who attempt to see
these identifications as tests of the theory, and therefore find
them to be tautological.
Were it the case, that we _identified_ relatively fit variants by
their mere persistence, and then were unable to find any attribute which
may have led to their persitence, we would have a refutation of the
theory. If the theory were consistently refuted, then fitness would be
seen as an empty concept and the aphorism would devolve to tautological
meaninglessness.
Are those who are posting in this thread proposing that there are
no demonstrated reasons why particular variants have survived; that we
know nothing about adaptation?
SotF was meant to be a summary of D's theory. The words may be
quite ambiguous when one considers all of their potential meanings, and
thus it may be a rather inefficinet summary. But I dont think the phrase
contradicts D's theory, if the words are understood in the ways that they
were meant to be understood. A question for all of you "tautologists": Do
you contend that darwinism is tautological? (if so, forget about SotF,
and lets discuss the theory in its formal explication). Or do you merely
find the phrase "SotF" to be a tautology while D's theory is not (in
which case..1) who cares? 2) I think you are supporting this assertion
merely by using the words in ways they were not intended to be used by
those who coined the phrase; 3) lets hear your attempts to summarize the
theory using unambiguous terms :) ).
Tom
Dr. David Rosen
the question of "tautology."
Example: I am a laser physicist. I work with laser cavities,
where photons are "self-replicating" entities. Photons are classified
in what we call cavity modes (mathematicians call them basis functions).
Feed-back loops, with an amplifying medium, gives these photons the
ability to "reproduce."
Before the laser is turned on, or before it has reached
equiblibrium, the photons are almost uniformly distributed in most
of the "modes." However, due to competition between photon modes for
energy, and due to damping forces, only a few "modes" survive till
equilibrium. Usually, only one mode survives. This mode tends to be
one that reduces the damping forces for itself. As an example, if there
is a scratch on the mirror, the surviving mode tends to be one that doesn't
shine on that scratch. However, the surviving modes and their proportions
are not unique. A slight disturbance can result in a whole new set of
survivoring modes. The concept of "fitness" seems, to me, analogous to
a cavity mode's Q-value. This Q-value is not defined by "survival",
although it obviously has a big effect of "survival." Q-value just
characterizes the damping values (i.e., the friction).
Analogous scenarious can be described for ANY system involving
a feed back loop. That whistling you hear when a loudspeaker has
unintentional feedback is the result of "natural selection," for
example. The origin of the first sound to enter the microphone is
of historical interest only. One moves the microphone, because the
fundamental process is well known. The initial sound is not logically
necessary for either an audio engineer to do his work, or even for
philosophical completeness.
Is laser engineering a tautology? Is loudspeaker repair?
David Rosen
dro...@inxs.chem.duke.edu
Peter Nyikos
>n>wee...@sagi.wistar.upenn.edu (Matthew P Wiener) writes:
>>>In article <4ch4un$2...@nntp5.u.washington.edu>, nyikos@math (Peter Nyikos) writes:
>>>>Reply-To: wee...@sagi.wistar.upenn.edu (Matthew P Wiener)
>>>>>It is a priori *impossible* for "survival of the fittest" to be a
>>>>>tautology, no matter how compelling one makes the argument.
>>>>I'd say that depends strongly on how one defines "survival" and
>>>>"fittest", as I pointed out in my follow-up to Alexey Merz.
And Alexey, while imperfectly handling "fittest" as meaning
"having a relative reproductive advantage", left the "survival"
side of the equation untouched.
>>>As long as you define them with reference to the real world, then it
>>>is not a tautology.
>> Species A was more fit at a given real point of time, t,
>> than Species B, because Species A went on for at least
>> one more generation, and Species B did not. IN OTHER
>> WORDS, Species A survived and Species B did not.
[Wiener deleted a bunch here.]
>>I disagree, see above. For "Species A" one can (for instance) put
>>*Homo sapiens*, for Species B put the passenger pigeon, and t can be
>>a certain point around the previous turn of the century which I can
>>look up if anyone is interested. Each relevant to which was "fit" at
>>time t, and FOR THE SAME REASON, which survived, if one uses the
>>"right" definitions for making SOTF a tautology.
>_Obviously_ if you give "fit" and "survive" the same definition, you
>have a tautology. Try giving semirealistic definitions.
I already gave one set, and you deleted them without even
commenting on whether you considered them realistic or not. Anyway,
I'd be more interested in learning what definitions biologists use.
>>>>>It is logically possible that, for unconsidered reasons having nothing
>>>>>to do with "tooth and claw", the not-so-fit will also survive.
>>Note that the highly reductionist definition of "fitness" up
>>there makes NO mention of "tooth and claw". Is this definition
>>of fitness the one you wish to adopt too? The biology books I've
>>seen which define fitness all seem to endorse it.
You didn't answer my question, Matt. As you can see from what
follows, you are slowing down discussion thereby.
>Try defining fitness locally--per creature, one generation--and survival
>via the long term--per species, say over thousands of generations.
>As is, I've never seen fitness defined as merely surviving to reproduce.
Try re-reading Maynard Smith's _Evolutionary Genetics_, for example,
where he defines the concept of fitness. Also note how he says
fitness is not supposed to be a property of individuals.
Try re-reading what Alex Merz wrote about
fitness = relative reproductive success
HE claimed it was unambiguous; from your comments I take it
you disagree with him, and agree with me that there is a good
bit of ambiguity in it.
>>Once you adopt it, one can simply define the statement
>>"the not-so-fit will also survive" intoa self-contradiction:
>Huh? No such self-contradiction is derived.
What makes you so confident of this, before even seeing
the definitions?
Let's say "not so fit" means that the mean (more precisely,
the geometric mean) population ratio from one generation to the next
is .99 or less. As one keeps multiplying this ratio from one generation
to the next, the numbers keep dropping until they get down to 1 individual,
at which point the species becomes extinct on the death of this
individual.
Don't blame me--YOU were the one who talked about "thousands of
generations" being the way to go with the word "survive".
>>>>That depends on where you draw the line between the fit and
>>>>not-so-fit. Also on where you draw the lines between "fit",
>>>>"fitter", and "fittest".
>>>First off, I do not interpret SOTF as a logical proposition, but an
>>>attempt to say something non-trivial about the real world.
>>Laymen often interpret SOTF a lot more broadly than the definition
>>"relative reproductive success", as suggested by your own use
>>of "tooth and claw". [...]
>Eh? My own use was only suggestive.
Suggestive of what? Thin air? Chopped liver? ;-)
I take it, since you deleted my question, that you do NOT want
to go the layman's route yourself. So, does this mean you
agree with Alexey Merz on "fitness" meaning relative reproductive
success?
And if so, what's your definition of "reproductive success"?
Does it mean the number of offspring that make it to reproductive
age, or does it mean the number that actually reproduce?
Remainder saved for later.
Peter Nyikos
>Peter Nyikos wrote:
>>I pointed out some ambiguities and have yet to see Alex or
>>anyone else address them.
>I want to applaud Peter Nyikos for his crisp understanding of the
>essence of this thread. I will try to elaborate on the ambiguities
>with the concept of fitness which he has correctly pointed out.
Thank you. Thank you.
>The general confusion and ambiguities concerning fitness is due to
>a whole army of hidden assumptions in biology. Scientists can afford
>to neglect these conditions because they just "work". I will try to
>summarize their most important aspects in two statements.
>--- An organism is fit by default
>But what the heck is an organism!? The answer to this question is
>everything but trivial. In fact, a thorough answer will be a _theory_
>of the organism. Basically an organism is a survival machine. It's
>intrinsic organization is that of _perpetuation_.
I gather from what you say below that "perpetuation" is not
synonymous with "reproduction":
So of all possible
>physical systems the organism belongs to the small class of systems
>that embodies perpetual motion (*). But note that this from the very
>start gives the biologist a lot of fitness for free. An organism is
>by definition a machine that tends to survive, and therefore has a
>high degree of fitness by default, before any formal concept of fitness
>has been defined.
Here you have definitely moved beyond "perpetuation" in the narrow sense
of "reproduction". But you rein us in with:
> But science rarely leaves any room for broad concepts. In order
>to be the subject of mathematical models the organism needs to be
>precisely defined which is done by operationally defining the organism
>as a _replicator_. The technical definition of fitness is therefore a
>function of replication.
Still awaiting a formal definition...
>--- Fitness is a function of measurement
[...]
>End of detour. Back to evolution. Suppose now that we did the same thought
>experiment for the measurement of fitness. Let us define a similar
>window parameter, W, as the number of generations over which we measure
>fitness. Let us now conceptually "tune" W from its minimal value (one
>generation) up to its maximum value (infinitely many generations) and see
>how it affects fitness. Starting with W=1 we will see a lot of
>fluctuations. That is, the resulting signal will be fuzzy. If we increase
>W to 10 we will filter out most of the high frequent noise and we will see
>the more low-frequent change which evolutionists refer to as "adaption" or
>micro-evolution. If we increase W to 10,000 then we will weed out much of
>this micro-evolution. For instance, the famous example of the increase in
>the black pepper moth population versus the white moth population would not
>turn up in the recording process at W=10,000. From this we can conclude that
>much of what is called evolution today is in fact species _homeostasis_. For
>low values of W we see the adaption, at high values of W we see the stasis.
>At even higher values of W we start recording extinction and speciation
>dynamics etc. etc.
> Alex Merz wrote that "in evolutionary theory fitness is
>_un_ambiguously defined as relative reproductive success." I hope I have
>demonstrated that this is a naive conception of fitness.
I agree, and I and wonder whether Alex was relying on some rather
naive (or, rather, oversimplified for the sake of the layman)
sources for his claim.
Fitness is a
>compound concept, governed by numerous parameters.
In what remained, you talked about the parameters including
a catchall "10% other factors". What we are still missing is
a definition of fitness that would give us a good handle on
what those factors might be.
Herb Huston
Benjamin H. Diebold <bdie...@minerva.cis.yale.edu> wrote:
}Anyway, Stephen J. Gould has a nice article on exactly this topic. It
}first appeared in Harper's Magazine in 1976, in response to an article by
}a creationist who claimed Darwinism was dead because it's a tautology.
}The response was expanded by Gould into an article, and published as
}chapter 8 of Laurie Godfrey's _Scientists confront creationism_ (1983).
If you're referring to Gould's "Darwin's Untimely Burial," which was a
response to Tom Bethell's "Darwin's Mistake," which was printed in the
February 1976 of _Harpers_, Gould's essay appears in _Ever Since Darwin_.
--
-- Herb Huston
-- hus...@access.digex.net
Onar Aam
>It seems to me an empty exercise; what is relevant is how the word is
>actually being used in the specific example under discussion.
I had two intents with my exercise. 1) to show that our measurements very much
rely on how we do them. In other words, you can to some degree decide what you
get out of the system. 2) the fitness concept will always be "slippery" (a lay
term for tautological) due to A) it's empirically circular definition and B) the
circular nature of the system under scrutiny.
>It is only those who
>wish to question the theory (mainly creationists) who attempt to see
>these identifications as tests of the theory, and therefore find
>them to be tautological.
I am neither a creationist nor trying to "test" the theory. I would say
that today's concept of natural selection and adaption in evolution is
directly comparable to the concept of "conditioning" and "adaptation" in
behaviorist psychology. Not only are they measured in the same way, they
are also 1D, surface concepts. Both are obviously correct (but not
necessarily complete). Typically, behaviorists claim that conditioning
can account for all learning in human development while evolutionists
claim that natural selection can account for all evolution. All in all
the two theories are very similar. So similar in fact that they belong
to the same paradigm of science.
Now, just like I have great respect for the results of behaviorism
I have tremendous respect for the findings of neo-darwinism. This, however,
does not stop me from challenging the basis for evolutionary theory. I
do not pursue behaviorism because it gives too little credit to Mind.
Similarly I don't pursue natural selection because it leaves the organism
out of the equation.
>A question for all of you "tautologists": Do you contend that darwinism is
>tautological?
Sure, it's inevitably tautological, as are most other scientific theories. So?
Onar.
Alex Merz
Thomas K. Dibenedetto <td...@umich.edu> wrote:
>Onar Aam (on...@hsr.no) wrote:
>
>[much ado about fitness]
Many thanks to Mr. Dibenedetto for his thoughtful reply. He said
what I would have - but said it better and more clearly.
-Alexey
Marc Robinson
>[...]Typically, behaviorists claim that conditioning
>can account for all learning in human development while evolutionists
>claim that natural selection can account for all evolution. All in all
>the two theories are very similar. So similar in fact that they belong
>to the same paradigm of science.
> Now, just like I have great respect for the results of behaviorism
>I have tremendous respect for the findings of neo-darwinism. This, however,
"evolutionists" do not claim that natural selection can account for all
evolution : neutral drift, mutation bias, etc, are taken into account.
So in challenging neo-darwinism you are not, in my opinion, challenging
"the basis for evolutionary theory".
____________________________________________________________
Marc Robinson
URA 2055 "Biometrie, Genetique, et Biologie des Populations"
Universite Claude Bernard Lyon 1
43, boulevard du 11 novembre 1918
69622 Villeurbanne cedex
France
tel : 72 44 80 00 - poste 41 43
e-mail : robi...@biomserv.univ-lyon1.fr
George W. Gilchrist
>...Typically, behaviorists claim that conditioning
>can account for all learning in human development while evolutionists
>claim that natural selection can account for all evolution.
No evolutionary biologist would make the claim that natural selection
can account for all evolution. This sort of fiction flows out of
simplistic popular and creationist versions of evolutionary theory.
The roles of genetic drift, meiotic drive, mutation, developmental
constraint, gene conversion, mass extinction, etc. may not be fully
understood, but they are certainly acknowledged as signficant factors
in evolution. Any college textbook on evolution will cover these,
should you wish to find our what evolutionary theory really is about.
Cheers, George
========================================================
George W. Gilchrist University of Washington
gil...@zoology.washington.edu Department of Zoology
Phone:(206)543-4859 Box 351800
FAX:(206)543-3041 Seattle, WA 98l95-1800
Onar Aam
>>can account for all learning in human development while evolutionists
>>claim that natural selection can account for all evolution.
>
>No evolutionary biologist would make the claim that natural selection
>can account for all evolution. This sort of fiction flows out of
>simplistic popular and creationist versions of evolutionary theory.
>The roles of genetic drift, meiotic drive, mutation, developmental
>constraint, gene conversion, mass extinction, etc. may not be fully
>understood, but they are certainly acknowledged as signficant factors
>in evolution.
I don't see the point here. Do you mean to say that biologists think
of mutations, genetic drift etc. somehow opposed to natural selection!?
Surely not.
Onar.
Grant B. Harris
explain. How does 'evolution' get us from non-replicating entities to
replicating entities? That, it seems to me, is the question.
-- Grant Harris
Shane McKee
(Somebody wrote...)
>>No evolutionary biologist would make the claim that natural selection
>>can account for all evolution. This sort of fiction flows out of
>>simplistic popular and creationist versions of evolutionary theory.
>>The roles of genetic drift, meiotic drive, mutation, developmental
>>constraint, gene conversion, mass extinction, etc. may not be fully
>>understood, but they are certainly acknowledged as signficant factors
>>in evolution.
>
>I don't see the point here. Do you mean to say that biologists think
>of mutations, genetic drift etc. somehow opposed to natural selection!?
>Surely not.
The previous poster is not saying that at all. Natural selection
acts on many characteristics of the phenotype determined in large
measure by the genome. To many aspects, however, NS can simply
shrug its metaphorical shoulders and say "not my problem".
Evolution in these cases can proceed largely by drift.
There are many reasons why things do not stay the same. Natural
selection accounts for most, if not all adaptations while simple
genetic mechanics, in the absence of any selection bias, accounts
for the variations present in a population. Natural selection is
not a mutagen.
Shane
Dr Shane McKee __ "As a matter of fact, I do have
Royal Hospitals / \ a plan - and it's so cunning,
====Belfast==== / 0 ] you could put a tail on it
=N.Ireland, UK= \/\_/ and call it a weasel."
Thomas K. Dibenedetto
: >A question for all of you "tautologists": Do you contend that darwinism is
: >tautological?
: Sure, it's inevitably tautological, as are most other scientific theories. So?
Tautology, like so many other concepts, is subject to extensive
philosophical analysis and debate regarding its meaning and
applicability. You are clearly using the term in a way which may have
profound philosophical justification (but so do other interpretations),
but I think it clear that you are _not_ using the term in the sense that it
has been used over the past century in debates over _this issue_. If you
bring a new, or different, meaning of the term to this debate (most
scientists do not consider most scientific theories to be tautologies),
you are simply sowing enormous confusion. WHy not make your case at the
appropriate level? Try to convince us what "tautology" really means, show
us how most scientific theories are tautologies; if you succeed at this,
then all of us will be able, on our own, to deduce that darwinism might
well be tautological as well.
But this is not what the "tautology" debate about darwinism has
been all about, and is not, I suspect, the level at which the original
question in this thread was aimed. Those who have made the tautology
charge against darwinism were not trying to make the case that it is just
like any other scientific theory; quite the contrary, they wished to
distinguish it from scientific theories to support the claim that it
should be rejected by thinking people. That is what the term means in
this context (the meaning is in the usage, right?).
--
Tom diBenedetto
Fish Division
University of Michigan Museum of Zoology
Sebastian Toombs
If something from the environment actively caused offspring to
develop increased fitness--?
If survival was not primarily related to an individuals
characterists, but was based on external or random factors-->
<--then Darwin's statement would not be true! Darwin stated that
the major force in the evolution of new species was that more
favorable traits are selectively passed on. This is an
empirically testable hypothesism, and not a tautology.
Alex Merz
>explain. How does 'evolution' get us from non-replicating entities to
>replicating entities? That, it seems to me, is the question.
Darwin's theory, the topic of this thread, does not address this issue.
Should Newton's theory specify where we can obtain spheres that are
perfectly elastic?
-Alexey
George W. Gilchrist
>I don't see the point here. Do you mean to say that biologists think
>of mutations, genetic drift etc. somehow opposed to natural selection!?
>Surely not.
Onar,
Genetic drift can certainly take a small population in a direction
opposite from that favored by natural selection. Likewise mutation,
gene conversion, meiotic drive, etc. In fact, most of these are
recognized in nature by their opposition to natural selection.
Evolutionary theory is more complicated that it seems. Better get a
good textbook or take some biology classes.
Epigene
The modern evolutionary synthesis is not a tautology. The expression
"survival of the fittest" can be defined in such a way that it is
tautological. This has no bearing on the "truth" or "falsehood" of the
expression. As a matter of fact, if formulated tautologically, it is
undeniably true the same way that, say, the distributive property is by
definition both tautological and true.
On the other hand, the modern evolutionary synthesis's truth or falsehood
is, one thinks, the object of a subtle attack attempting to reject it on
logical grounds. These attempts are illogical, however, themselves for
two reasons: first, because the formulations are abstractions that do not
cast the claims of evolutionary theory in biological terms; second,
because evolution has already been proven to have occurred. It has been
painstakingly documented by a number of researchers in different species
IN OUR LIFETIME. Evolution and natural selection are no longer theories,
but verifiable facts that are central to the biological research program.
Peter Nyikos
> Clearly appeals to reason (postings in this thread which have
>pointed out the SotF is a popular summary of D's theory, rather than its
>formal explication) have fallen on deaf ears. If there is a fascination
>with analyzing this aphorism, so be it. Can we all agree that there are
>variants within populations? And that over the course of many
>generations, some of these variants persist, whereas others disappear?
>This is what "survival" refers to; not the survival of an individual, but
>the persistence of a variation. This differential persistance is posited
>to be non-random; to be correlated with a certain efficiency in
>utilizing the resources of the environment in which the variants find
>themselves. This efficiency is seen (in a theoretical sense) as a measure
>of adaptation. It is the idea behind the concept "fitness".
In the grand overall sense of the word, certainly. But some
books (and some individuals on this thread, particularly
Alex Merz) seem to insist on a reductionist meaning of the
term where "the proof of the pudding is in the eating" and
the only thing that enters into even defining "fitness" is the
end result of "relative reproductive success".
Was this the way Darwin looked on the word "fitness"?
> Attempts to measure this fitness are part of the grand research
>project of evolutionary biology. We can look at the raw persistance of the
>variants and _identify_ the distribution of relative fitness, because we are
>operating within the assumptions of the theory (the "survivors" are
>_expected_ to be the fit ones in the theory). Thus we tend to measure
>fitness by looking at reproductive success, for clearly this is the
>engine behind persistance.
Just how divided are biologists between your use of the word
and the reductionist sense?
Such identifications do not test the validity
>of the theory; they are not intended to.
Which theory? SOTF?
This exercise is part of
>_doing_ evo biology, not testing its assumptions. It is only those who
>wish to question the theory (mainly creationists) who attempt to see
>these identifications as tests of the theory, and therefore find
>them to be tautological.
Although I have been convinced of the reality of evolution
since the age of 12, I think tautology is a separate issue,
very much alive if one insists on the reductionist meaning
of the terms. I almost despair of getting Matt Wiener to
realize that this is what I keep driving at in my follow-ups
to him and Merz, or of getting Merz to pay any attention
to what I am saying. Do you, at least, see what my point is?
> Were it the case, that we _identified_ relatively fit variants by
>their mere persistence, and then were unable to find any attribute which
>may have led to their persitence, we would have a refutation of the
>theory.
The non-reductionist theory of SotF, I take it. I agree.
If the theory were consistently refuted, then fitness would be
>seen as an empty concept and the aphorism would devolve to tautological
>meaninglessness.
Indeed.
> Are those who are posting in this thread proposing that there are
>no demonstrated reasons why particular variants have survived; that we
>know nothing about adaptation?
I certainly am not. My quarrel is with the reductionist concept
and only the reductionist concept of "fitness". I've seen a
popular book on genetics which insists that it is a popular
misconception that the meaning of fitness is anything MORE
than this reductionist concept. So I keep wondering which
side most professionals are on.
[I can try to look up the book again, if you wish.]
Peter Nyikos -- standard disclaimer --
Peter Nyikos
me...@ohsu.EDU (Alex Merz) writes:
I suspect you may have misread him. I've tried to post a
follow-up to him which pointedly refers to you several times.
If it's rejected by the moderators for any reason, I'll
try to repost it without the pointed references.
Onar Aam
Tom diBenedetto writes:
>Tautology, like so many other concepts, is subject to extensive
>philosophical analysis and debate regarding its meaning and
>applicability. You are clearly using the term in a way which may have
>profound philosophical justification (but so do other interpretations),
>but I think it clear that you are _not_ using the term in the sense that it
>has been used over the past century in debates over _this issue_.
I think this is true, but only to a certain extent. My use of the term may
be different in form, but its _essence_ is the same. That is, when the term
tautology was first coined its intent was to capture the idea of
self-justification. Only later was it given a formal definition. But formal
definitions only captures so and so much of an idea. This is true for all
conceptual framework, even formal logic. Logic existed long before formal
logic as the emergent structure of thinking. When I use the term tautology,
I use it in a broad sense, including its current formal definition.
>Try to convince us what "tautology" really means, show
>us how most scientific theories are tautologies;
Here I refer to the _formal_ definition of a tautology from mathematics. Suppose
we have n propositions. Then a tautology can be expressed in the form
(p1^p2^p3^...^pn) => q
All this says that q is true if and only if p1,p2,... and pn are true. Now, all
of mathematics is built on this basic structure meaning that it is all a gigantic
tautology. In this meaning of the term tautology simply means _consistency_.
Clearly, then, most scientific theories are formal tautologies.
But this is not what is meant when tautology is discussed in relation to
evolutionary theory. Here tautology is used exclusively as a negative concept
meaning self-justification or non-informativity. Since we now have entered
the empirical realm we can no longer speak of a binary tautology, but rather
_fuzzy_ tautologies. The meaning of this term is the answer to the question
"how easy is it to construct consistent explanations with theory X?" The easier,
the more tautological X is. Natural selection has as a high tautological value
because it's so easy to construct consistent working explanations to evolutionary
phenomena. Let me give you a (fictious) example from memetics. (Memes are
information viruses/genes that are subject to natural selection) We start with an
observation: everyone in the world consumes food. We want to argue that this
widespread behavioral pattern is an instance of a successful meme and posit a
meme which we call the food-meme, which is essentially the belief that food is
vital for our survival. We may posit that this food-meme arised in symbiosis with
the economic systems. In economy we are familiar with several different
industries: primary, secondary and tertiary. Tertiary industry builds on
secondary which builds on primary industry (i.e. food production). Therefore, if
the basis for primary industry (need for food) were to collapse then the entire
global economy would collapse with it. It is therefore vital for an economy to
sustain the need for food, _even if there were no physiological basis for it_.
This is where the food-meme enters the picture. We may posit that economy is
sustained precisely because of the success of the food-meme. We are familiar with
the power of cultural indoctrination during childhood, and we observe that food
is indeed indoctrinated in people throughout their childhood. ("you have to eat
to grow") So much that they sustain this cultural trait in adulthood. There are
also self-amplifying psychological mechanisms at work here, e.g. the inverse
placebo effect. Because people believe they will become hungry if they don't eat,
they _will_ become hungry. This belief is so strong that long periods without
food will have psychosomatic effects. In some cases the post-traumatic effects of
this may even be lethal. People may in fact "starve" to death. All this explains
the great reproductive success of the food-meme.
Onar.
Grant B. Harris
>You're confusing evolution with abiogenesis...
No, I'm not, as I have never heard of abiogenesis. Yes, we are argueing
semantics, but semanitics both constrain and enable how we think and see
and act. The context in which I am using the word 'evolution' is
'entropy and evolution'. If you wish to exclude 'abiogenisis' from the
realm of dialogue regarding 'evolution', I guess we don't have much to
argue about.
-- Grant Harris
Peter Nyikos
adding some of mine. :-)
on...@hsr.no (Onar Aam) writes:
>Tom diBenedetto writes:
>>Tautology, like so many other concepts, is subject to extensive
>>philosophical analysis and debate regarding its meaning and
>>applicability. You are clearly using the term in a way which may have
>>profound philosophical justification (but so do other interpretations),
>>but I think it clear that you are _not_ using the term in the sense that it
>>has been used over the past century in debates over _this issue_.
>>Try to convince us what "tautology" really means, show
>>us how most scientific theories are tautologies;
>Here I refer to the _formal_ definition of a tautology from mathematics. Suppose
>we have n propositions. Then a tautology can be expressed in the form
> (p1^p2^p3^...^pn) => q
Sorry, Onan, I feel impelled to interrupt here:
You used the wrong symbol: you should have put <=>
>All this says that q is true if and only if p1,p2,... and pn are true. Now, all
>of mathematics is built on this basic structure meaning that it is all a gigantic
>tautology.
Incorrect. Had you talked only of => you would have had a
strong point.
We start from a basic set of axioms and derive everything else from
them in mathematics, but there are many different sets of axioms
one can use as one's starting point and one does not get both
necessary and sufficient conditions in the majority of one's theorems.
>Clearly, then, most scientific theories are formal tautologies.
Non sequitur, I'm afraid. Most scientific theories are not
reducible to pure mathematics AFAIK.
I would, however, go along with any definition of "tautological"
that has to do with starting from a set of empirical data
AND various definitions, and deriving a conclusion without ever
inputting any further empirical data. So "survival of the
fittest" could have its component "survival" defined with reference
to thousands of generations, while "fitness" to only one, and yet it
could be tautological even without having to bring in ANY observations
from the real world, simply by straightforward mathematical
reasoning if the definitions were well chosen.
It's not the mathematical meaning of "tautological", but I
don't mind.
Shane McKee
FORWARDED MAIL -------
From: sh...@reservoir.win-uk.net (Shane McKee)
Date: 20 Jan 96
Originally To: g...@workframe.com
Grant Harris wrote:
Grant,
If you're talking about "entropy and evolution" I think it's fairly
clear that during the majority of Earth's history, evolution has
been proceeding nicely down an entropically favourable path. The
getting-from-non-life-to-life only had to happen once, and current
experiments on molecular systems are revealing tantalizing
glimpses of what processes might be involved. The vast majority of
evolution has had zip to do with this event, other than following
from it, and I feel that using the term "Evolution" as a tag for
it is a tad misleading. We all sometimes need to be more specific as
to what we're talking about.
"Abiogenesis" is a word which is generally accepted to mean
"life-from-non-life", and is probably more appropriate. Someone
else may have other ideas.
Now what did entropy have to do with this discussion?
Matthew P Wiener
>I would, however, go along with any definition of "tautological" that
>has to do with starting from a set of empirical data AND various
>definitions, and deriving a conclusion without ever inputting any
>further empirical data. So "survival of the fittest" could have its
>component "survival" defined with reference to thousands of
>generations, while "fitness" to only one, and yet it could be
>tautological even without having to bring in ANY observations from
>the real world, simply by straightforward mathematical reasoning if
>the definitions were well chosen.
This isn't tautological at all.
It takes observation that all the hidden assumptions that go into
the proof are true about the real world.
For example, spontaneous abiogenesis is a priori possible.
Peter Nyikos
>In article <4dp9b9$l...@nntp5.u.washington.edu>, nyikos@math (Peter Nyikos) writes:
>>I would, however, go along with any definition of "tautological" that
>>has to do with starting from a set of empirical data AND various
>>definitions, and deriving a conclusion without ever inputting any
>>further empirical data. So "survival of the fittest" could have its
>>component "survival" defined with reference to thousands of
>>generations, while "fitness" to only one, and yet it could be
>>tautological even without having to bring in ANY observations from
>>the real world, simply by straightforward mathematical reasoning if
>>the definitions were well chosen.
>This isn't tautological at all.
I'd call this "fuzzy tautology" but I yield to Onar Aan on this
one. He seemed to be opting for an even more sweeping use of
the word than the above.
>It takes observation that all the hidden assumptions that go into
>the proof are true about the real world.
Trivially true and irrelevant to the tautologicality.
In the SotF context, I've already given you a concrete
example which you have yet to address. It proceeds from
premises to conclusion without ANY input from the "real world",
in the way I depicted above, only it was about the flip
side of the coin, "the not-so-fit do not survive".
And if the real world doesn't contain any provable examples
of not-so-fit individuals, that still does not invalidate
the reasoning nor the conclusion.
>For example, spontaneous abiogenesis is a priori possible.
This has nothing to do with survival of the fittest,
which applies only after life has appeared, and
it is irrelevant whether it appeared by spontaneous
abiogenesis, spores from another solar system, or
divine creation.
Did Darwin ever concern himself with abiogenesis, by the way?
>-Matthew P Wiener (wee...@sagi.wistar.upenn.edu)
Joachim Dagg
simplistic statements:
on...@hsr.no (Onar Aam) writes:
>DiBenedetto wrote:
>>A question for all of you "tautologists": Do you contend that darwinism is
>>tautological?
>Sure, it's inevitably tautological, as are most other scientific theories. So?
1. When scientists seek for theories they seek for something that fits on
all natural conditions.
2. A statement that is allways true is a tautology.
If one sees all natural conditions as all conditions considerable,
theories are tautologies, so what? If one sees still more conditions
considerable than the ones naturally possible, scientific theories are
no tautologies. That are the two poles I see. Onar seems to consider
more than the natural conditions for a system and jet find scintific
theories about them tautological. I keep thinking about that!
Joachim
Easy:
Your heart beats slow, your mind is free, your arms are hanging besides your
body - I repeat - your heart beats slow, your mind is free, your arms
are hanging besides your body - I repeat - your heart beats slow ...
Onar Aam
>>All this says that q is true if and only if p1,p2,... and pn are true. Now, all
>>of mathematics is built on this basic structure meaning that it is all a gigantic
>>tautology.
>
>Incorrect. Had you talked only of => you would have had a
>strong point.
Oops, you are of course absolutely correct. This makes the argument more
complex, but I think it still holds. The structure we have to look at is
not the formal tautology in itself, but the deduction process and view it
as an iterative feedback structure, much like a growing crystal. Clearly
this growth is constrained by the structure of the crystal itself, and
mathematics must thus be self-contained.
Such an evolutionary perspective on mathematics produces a panoply
of questions: is it self-similar? Can there co-exist several independent
axiomatic systems that will never ever grow into each other? If no, can we
determine the degree of inter-webbedness between different axiomatic
systems? My own answers to these questions are: yes, no and maybe.
>I would, however, go along with any definition of "tautological"
>that has to do with starting from a set of empirical data
>AND various definitions, and deriving a conclusion without ever
>inputting any further empirical data. So "survival of the
>fittest" could have its component "survival" defined with reference
>to thousands of generations, while "fitness" to only one, and yet it
>could be tautological even without having to bring in ANY observations
>from the real world, simply by straightforward mathematical
>reasoning if the definitions were well chosen.
This is not my view of fitness (or scientific theories in general) as
tautologies. The key here is that scientific theories, like mathematics,
are self-contained and self-sustaining to some degree. A successful
theory acts like a filter on reality. It focuses on some very specific
aspects of reality and ignores most others. In this way a good theory will
always to some degree be self-confirming; it only looks for the things that
it predicts, not at the things it _doesn't_ predict. For instance, physics
is the most successful branch of science throughout history, even so it is
riddled with paradoxes and problems that rarely surface. *IF* they had
surfaced more often then physics would have been in a lot more trouble than
is the case.
Onar.
Peter Nyikos
>>>It takes observation that all the hidden assumptions that go into
>>>the proof are true about the real world.
I misunderstood this, and apologize for the confusion. The hidden
assumptions that go into the proof may or may not require
observation; it all depends on what those assumptions are.
I gave a concrete example a while back but see now that I
have to add some assumptions to make it hold:
>> It proceeds from premises to conclusion
>>without ANY input from the "real world", in the way I depicted above,
>>only it was about the flip side of the coin, "the not-so-fit do not
>>survive".
>And only observation will determine if the real world in fact follows
>that model.
Not if I modify the model in the appropriate way. Let's recapitulate
what I did a while back:
fit = having a relative reproductive advantage
survival = having descendents over a significant [and we
could haggle over how many is "significant"] number of generations
fittest = those whose relative reproductive advantage passes
a certain given point.
Now, we did some haggling about "significant", and ended up
opting for "thousands of generations". Also I adopted
the Maynard Smith definition for "fit" in order to give
added precision to the above loose definition:
`fitness of an A zygote' = `probability that an A zygote will
survive to breed' x `expected number of offspring, given that
it does survive'.
--John Maynard Smith, _Evolutionary Genetics_,
1989, Oxford University Press, p. 37
My demonstration that at least "the not-so-fit do not survive"
can be a tautology, went as follows:
Let's say "not so fit" means that the mean (more precisely,
the geometric mean) population ratio from one generation to the next
is .99 or less. As one keeps multiplying this ratio from one generation
to the next, the numbers keep dropping until they get down to 1 individual,
at which point the species becomes extinct on the death of this
individual.
Wiener found a loophole in the argument, and I patch it
up below by shifting from "species" to a certain definition
of "type".
If you use ".9 or less" rather than ".99 or less," then 500
generations would already be enough to wipe out a "type" numbering
10 quadrillion at the start. [American definition of 1 quadrillion
equals one thousand million million.]
In the sense used here,
type = "the set of members of the species
that exist at a given time T, and their descendants".
Species that do a
>pathetic job of reproducing, but have the necessary abiogenetic help
>now and then, will also survive, despite SOTF predicting their doom
>with mathematical certainty.
No, it's the flip side, "non-survival of the not-so-fit" that
would be predicting it.
But what about SotF? replenishment "from below" only makes
an already fit species more likely to survive. So you would
have to factor in the flip side, the possibility of members
of the species having their descendants becoming members of
a new species, if you want to make SotF escape tautology.
But we can take care of that by simply disqualifying those
descendants from contributing to the definition of "fitness"
above.
Epigene
The question is, has natural selection been observed to occur in nature?
The answer is "yes." The fact that some aspects of the theory can be
defined as tautologies based on a narrow definition of "fitness" is no
longer a relevant consideration. Natural selection and evolution are
observable facts, thanks to the work of people like Peter and Rosemary
Grant. This tautological idea, which Dawkins rightly called "a whimsical
little conceit" (The Extended Phenotype) should be ignored by
well-educated people.
EPIGENE (Scott Hatfield)
Matthew P Wiener
>wee...@sagi.wistar.upenn.edu (Matthew P Wiener) writes:
>>>>It takes observation that all the hidden assumptions that go into
>>>>the proof are true about the real world.
>I misunderstood this, and apologize for the confusion. The hidden
>assumptions that go into the proof may or may not require
>observation; it all depends on what those assumptions are.
Lack of spontaneous generation requires observation.
>>Species that do a pathetic job of reproducing, but have the
>>necessary abiogenetic help now and then, will also survive, despite
>>SOTF predicting their doom with mathematical certainty.
>No, it's the flip side, "non-survival of the not-so-fit" that
>would be predicting it.
That's part of SOTF, as I originally explained. If everyone survives,
SOTF is vacuously true and pointless. If there's a culling--which is
what is of interest--then SOTF tells us who loses.
Peter Nyikos
>In article <4eod1g$2...@nntp5.u.washington.edu>, nyikos@math (Peter Nyikos) writes:
>>wee...@sagi.wistar.upenn.edu (Matthew P Wiener) writes:
About various "proofs" that Survival of the Fittest, SOTF, is
a tautology:
[Wiener:]
>>>>>It takes observation that all the hidden assumptions that go into
>>>>>the proof are true about the real world.
[Nyikos:]
>>I misunderstood this, and apologize for the confusion. The hidden
>>assumptions that go into the proof may or may not require
>>observation; it all depends on what those assumptions are.
[Wiener, earlier:]
>>>Species that do a pathetic job of reproducing, but have the
>>>necessary abiogenetic help now and then, will also survive, despite
>>>SOTF predicting their doom with mathematical certainty.
>>No, it's the flip side, "non-survival of the not-so-fit" that
>>would be predicting it.
>That's part of SOTF, as I originally explained.
Onar Aam disagreed, at least at one point; did you two
patch up your disagreement?
Now, the clue about the SOTF is not what it says but what it _doesn't_
say, namely that those that are unfit will tend NOT to survive.
--<4crk6l$9...@nntp5.u.washington.edu>
If everyone survives,
>SOTF is vacuously true and pointless. If there's a culling--which is
>what is of interest--then SOTF tells us who loses.
False dichotomy. SOTF could still be vacuously true and
pointless in the latter case, once one makes the rather
trivial observation that abiogenesis is no longer taking
place, as I showed in the article to which you are following
up. This goes back to what I suggested could reasonably
be called a "fuzzy tautology" (although Onar Aam would include
a lot of ADDITIONAL science under this rubric):
I would, however, go along with any definition of "tautological" that
has to do with starting from a set of empirical data AND various
definitions, and deriving a conclusion without ever inputting any
further empirical data. So "survival of the fittest" could have its
component "survival" defined with reference to thousands of
generations, while "fitness" to only one, and yet it could be
tautological even without having to bring in ANY observations from
the real world, simply by straightforward mathematical reasoning if
the definitions were well chosen.
I actually fixed up the definitions to eliminate abiogenesis
from the factors entering into SOTF, but one could also go
with my first sentence, and, after factoring in the observations
that abiogenesis is not taking place, nor is another species
evolving continuously into the one under study, proceed by
straightforward calculations without ever inputting any
further data.
Onar Aam
>[Wiener, earlier:]
>>>>Species that do a pathetic job of reproducing, but have the
>>>>necessary abiogenetic help now and then, will also survive, despite
>>>>SOTF predicting their doom with mathematical certainty.
>
>>>No, it's the flip side, "non-survival of the not-so-fit" that
>>>would be predicting it.
>
>>That's part of SOTF, as I originally explained.
>
>Onar Aam disagreed, at least at one point; did you two
>patch up your disagreement?
>
> Now, the clue about the SOTF is not what it says but what it _doesn't_
> say, namely that those that are unfit will tend NOT to survive.
> --<4crk6l$9...@nntp5.u.washington.edu>
I see how that statement of mine may have caused confusion. What I meant is
that the flip side of SOTF is implied by it, but not _explicitly_ stated. As such
there is no disagreement between Wiener and me.
[Wiener stated:]
>> If everyone survives,
>>SOTF is vacuously true and pointless. If there's a culling--which is
>>what is of interest--then SOTF tells us who loses.
[to which Nyikos replied:]
>False dichotomy. SOTF could still be vacuously true and
>pointless in the latter case
Exactly my point. Fitness is _defined_ by those survivers. Therefore we could
equally much say that those that survive tells us something about the fittest.
In other words we have a circularity here between observation and the concept of
fitness.
Onar.
Matthew P Wiener
>>If everyone survives, SOTF is vacuously true and pointless. If
>>there's a culling--which is what is of interest--then SOTF tells us
>>who loses.
>False dichotomy. SOTF could still be vacuously true and pointless in
>the latter case, once one makes the rather trivial observation that
>abiogenesis is no longer taking place, as I showed in the article to
>which you are following up.
Which, as I said, is an empirical observation. Whether it is trivial
or not, it is still empirical.
Arlin Stoltzfus
actually want to conceive of natural selection as non-tautological, you
could just take not of the FACT of random genetic drift, which provides
an alternative to natural selection, and which demonstrates an alternative
outcome of sorting, something other than "the fittest survive".
If you don't understand this, then go and look up a paper by A. Buri
published in _Genetics_ about 30 YEARS AGO. Buri took the F1 of a large
cross of two different Drosophila lines homozygotic for alleles A1 or
A2 at a particular locus, and split it into many small populations.
Each small experimental population was bred separately for many
generations. Since all flies were A1/A2 heterozygotes at the start of
the experiment, the initial frequency for allele A1 was 0.5 in each
population. As the experiment progressed the allele frequencies
diffused away from this. When the experiment was stopped, most of
the populations were monomorphic, about half of these had lost A1 and
the other half had lost A2.
When the same alleles are competed in a very large population, one
of the alleles tends to do better. In Buri's experiment, either could
be lost. So how is fitness defined, Onar? Think about it.
Instead of getting into deep philosophical conversations about whether
natural selection is a tautology, why not accept the fact that there
must exist a conception of natural selection that is not a tautology?
Then you can start addressing more interesting questions, for instance,
exactly what formulation of natural selection is not a tautology? Why
does it appear to be tautologous when it really isn't?
Since Ernst Mayr's arguments apparently have not convinced you, I
would suggest the following:
Susan Mills and John Beatty, The Propensity Interpretation of Fitness
(anthologized in Elliot Sober's _Conceptual Issues in Evolutionary
Biology_).
E.S. Vrba and S.J. Gould. The hierarchical expansion of sorting and
selection: sorting and selection cannot be equated. Paleobiology
12(2), 217-228.
Arlin
Arlin Stoltzfus (ar...@is.dal.ca)
Department of Biochemistry
Dalhousie University
Halifax, NS B3H 4H7 CANADA
ph. 902-494-3569; fax 902-494-3569
Onar Aam
>actually want to conceive of natural selection as non-tautological, you
>could just take not of the FACT of random genetic drift, which provides
>an alternative to natural selection, and which demonstrates an alternative
>outcome of sorting, something other than "the fittest survive".
I think this is wrong. Genetic drift does not in any way contradict
natural selection. Drift can only take place in systems with low/neutral
selective pressure.
>When the same alleles are competed in a very large population, one
>of the alleles tends to do better. In Buri's experiment, either could
>be lost. So how is fitness defined, Onar?
As I've said many times already, fitness is a FUZZY concept (and
hence "survival of the fittest" is a fuzzy tautology). In Buri's
experiment the populations were small enough for the fuzz to play
a role. In large populations, however, this fuzz tends to be watered
out.
>Instead of getting into deep philosophical conversations about whether
>natural selection is a tautology, why not accept the fact that there
>must exist a conception of natural selection that is not a tautology?
This I do accept. I follow Ben Goertzel in his concept of fitness.
He takes an ecological/social view on fitness. In other words, he
reads fitness as in "fits like a glove". This ecological definition
of fitness is independent of survival. In this context the statement
"survival of the fittest" therefore becomes a lot more informative
and radical because it says something definite about the structure
of survival. Species that compete for the same resources will tend
to exhaust each other in the long run and thereby run the risk of
going extinct altogether. Species, however, that comply their own
niche to other niches by avoiding interference will have a greater
chance of survival because they avoid exhausting competition. In
other words, an ecology may be considered a structurally closed web.
Niches are nodes in the network. Each of these nodes may be viewed
as REPELLORS. A fit ecology may therefore be defined as a network
where the distance between all nodes is MAXIMIZED. (I.e.
interference between nodes is minimized) This is also related to the
classical mathematical problem of sphere packing in n-space. (e.g.
how to get the largest number of oranges into a box) In a mature
ecology the nodes will be packed together in a more efficient manner
than in young ecologies. In other words, mature ecologies have less
open space between the nodes. This makes the ecology as a whole
structurally more inpenetrable from the outside because there are
no free niches available.
This of course does not mean that old ecologies are
inpenetrable, (cf rabbits in Australia) or that an ecology evolves
towards uniformity (cf mass extinction patterns).
To read more about an ecological concept of fitness see Ben
Goertzel's "The Evolving Mind".
>Then you can start addressing more interesting questions, for instance,
>exactly what formulation of natural selection is not a tautology? Why
>does it appear to be tautologous when it really isn't?
In my view expelling the question of tautology to philosophy is a
way of dismissing a very fundamental aspect of survival from
science. That's a shame.
Onar.
Pelle Ingvarsson
has nothing to do with selction pressures. The mangnitude of the drift,
however, is determined by the size of the system (i.e. population size). If
population size is small enough, random genetic drift will override very strong
selection pressures. If the populations were small enough in Buri's experiment
(which I think they were), drift will be the main process that determines the
outcome (i.e. half of the populations fix for allele 1 and half for allele 2
by pure chance). In a large population (N>100, say) the effects of drift will
be very small and selection will have an effect. In this case the allele
favoured by selection will go to fixation. There is no contradiction bewteen
the results what so ever.
-Pelle
---------
P.K. Ingvarsson
Dept. of Ecological Botany
University of Umea
SWEDEN
Mark Barton
>Only if you insist on looking at it that way. On the other hand, if you
>actually want to conceive of natural selection as non-tautological, you
>could just take not of the FACT of random genetic drift, which provides
>an alternative to natural selection, and which demonstrates an alternative
>outcome of sorting, something other than "the fittest survive".
>
>If you don't understand this, then go and look up a paper by A. Buri
>published in _Genetics_ about 30 YEARS AGO. Buri took the F1 of a large
>cross of two different Drosophila lines homozygotic for alleles A1 or
>A2 at a particular locus, and split it into many small populations.
>Each small experimental population was bred separately for many
>generations. Since all flies were A1/A2 heterozygotes at the start of
>the experiment, the initial frequency for allele A1 was 0.5 in each
>population. As the experiment progressed the allele frequencies
>diffused away from this. When the experiment was stopped, most of
>the populations were monomorphic, about half of these had lost A1 and
>the other half had lost A2.
>
>When the same alleles are competed in a very large population, one
>of the alleles tends to do better. In Buri's experiment, either could
>be lost. So how is fitness defined, Onar? Think about it.
>
>Instead of getting into deep philosophical conversations about whether
>natural selection is a tautology, why not accept the fact that there
>must exist a conception of natural selection that is not a tautology?
>Then you can start addressing more interesting questions, for instance,
>exactly what formulation of natural selection is not a tautology? Why
>does it appear to be tautologous when it really isn't?
SotF is obviously a slogan, and to make sense of it we have to add fine
print. Possibly there are other viable interpretations, but I think it
makes most sense with the following fine-print: There is _differential_
survival of the fittest _traits_, _into the next generation_, where
"fittest" means leaving the most offspring _on average_. In this
interpretation I see it as fundamentally a claim about heredity. I think
it appears tautologous now because we have a fully worked-out and
separate
genetic theory of heredity in which most of the original content is just
plain obvious.
To see this consider an alternate model of heredity in which SotF is
_not_
true. Suppose that the structure and behaviour of an organism is
determined by three factors: (i) a Platonic ideal or Biblical kind, (ii)
fresh random variation (iii) inheritance of variation with a one or two
generation timeout. This model would reproduce at least superficially
what
was known about heredity when Darwin started, and something very like it
was presumably in serious consideration. If we apply a steady selection
pressure in this model, it will almost immediately saturate. There will
be
organisms with different traits having different numbers of offspring and
we can identify the fit traits accordingly. However although an organism
with a "fit" trait will by definition have more offspring, the trait will
not appear with increased frequency in those offspring because the
population
will come to an equilibrium near the underlying ideal form defined by
(i) and in balance between the rate at which the trait is generated by
(ii)
and the rate at which it expires by (iii). Thus the trait will not
"survive" differentially into the next generation and will not come to
dominate the population.
If we continue to focus on the _traits_, as opposed to the genes which we
now
know to underlie the traits, then SotF is still true and interesting and
we
can identify some reasons _why_ it is true:
(a) To a fair approximation genes represent the entire information
content of
the final organism. Although the environment has a big effect of course
it is
not a systematic one, and there is nothing like a Platonic ideal which
preserves a core of the design independently of the genome. This means
there is nothing to act as an anchor and stop evolution proceeding
arbitrarily far from a given starting point.
(b) There are no core genes which are absolutely protected from mutation
and natural selection. This also removes a possible anchor.
(c) Successful genes are approximately immortal - there is no systematic
effect
except natural selection which removes traits.
There are probably many more facts about genetics which guard against
ways
that SotF could in principle fail but doesn't and it would be instructive
to look for them. It is by no means obvious that SotF will always hold
and it
would also be instructive to look for situations in which it fails. In
fact the experiment by Buri could be interpreted as a minor failure of
SotF
depending on how you interpret the fineprint about "on average".
SotF becomes tautologous when you redefine fitness from being a property
of a
_trait_ to being a property of a _gene_. However this is the sensible
thing
to do once you have confirmed that heredity works largely by genes. As
Dawkins says, "It was only later that fitness was adopted as a technical
term. Biologists thought they needed a word for that hypothetical
quantity
that tends to be maximised as a result of natural selection. They could
have chosen 'selective potential', or 'survivability', or 'W' but in fact
they hit upon 'fitness'. They did the equivalent of recognizing that the
definition they were seeking must be 'whatever it takes to make the
survival of the fittest into a tautology'. They redefined fitness
accordingly." (The Extended Phenotype, p.181-2.) Darwin et al. couldn't
have intended SotF in this sense, because they didn't have the modern
genetic theory of heredity, and conversely, if you are going to dredge
the
slogan up in the late 20th century, it makes no sense to use the modern
gene-based definition of fitness.
Cheers,
Mark B.
Onar Aam
saying "those that tend to survive will survive".
> As
>Dawkins says, "It was only later that fitness was adopted as a technical
>term. Biologists thought they needed a word for that hypothetical
>quantity
>that tends to be maximised as a result of natural selection. They could
>have chosen 'selective potential', or 'survivability', or 'W' but in fact
>they hit upon 'fitness'.
Herein lies much of my point. The technical meaning of fitness is simply a single
coefficient. Much of this discussion has been about whether fitness is an
informative concept. I don't think that a concept that can be reduced to a single
variable is very informative.
Onar.