[email protected] (Larry Moran) wrote in message news:<
[email protected]>...
> On Mon, 2 Feb 2004 07:03:10 +0000 (UTC), Jim Menegay <
[email protected]> wrote:
> >
[email protected] (Larry Moran) wrote in message
> > news:<
[email protected]>...
First, let me say that it was not my intention to create and defend a taxonomy of the roles of definitions-of-
phenomena in epistemology. My main point was that definitions serve multiple purposes, and that
different definitions serve different purposes better than others.
Second, let me express my regret that I chose the words "loose" and "precise" for the two kinds
of definitions - I fear that Moran prefers to think of himself as "precise", rather than "loose",
and he is offended that I have made Edser's kind of definition "precise", whereas Moran's kind is
"loose". My apologies for that - it wasn't my intention. My view is that both kinds of
definitions are necessary, and perhaps additional kinds, as well. "Precise" is not better than
"loose" in this context.
>
> [snip]
>
> >> Please give me a precise definition of evolution that works and is a "creature of theory."
> >
> > "Evolution is change in gene frequencies caused by differential survival and reproduction of
> > organisms." This definition is a creature of the neo-Darwinist theory (Fisher, Wright, et al.)
> > It certainly wasn't Darwin's precise definition - he had never heard of genes. His definition
> > involved things like speciation and adaptation, even though his causal ideas were not that
> > different from Fisher's. People between Darwin and Fischer (Bateson, for example) also talked
> > about evolution without necessarily talking about changes in gene frequencies.
>
> Thank-you very much for offering a concrete example that we can examine. The definition of
> evolution that I prefer is the one in the article that I wrote for the Talk.Origins Archive. I
> don't claim that this definition is original. In that article I've tried to summarize the
> consensus among evolutionary biologists.
>
>
http://www.talkorigins.org/faqs/evolution-definition.html
>
> "Evolution is a process that results in heritable changes in a population spread over many
> generations."
>
I like this definition. "Heritable changes" is a big improvement over "change in gene frequencies".
"Spread over many generations" is also a desirable clarification, though I will quibble about one of
its side-effects below. Of course, this is an example of what I have called a "loose definition",
since it doesn't mention causality. It might also work as a precise definition - I don't insist that
a precise definition HAS TO talk about causality.
> There are many variations of this definition. Some of them are mentioned in the article. The most
> common variant refers to changes in the frequency of alleles in a population.
>
> Keep in mind that this is a minimal definition of evolution and doesn't address the birth and
> extinction of populations. It is, admittedly, a creature of prior knowledge in that it assumes an
> understanding of populations of organisms and of heredity.
The "minimal" nature of the definition tends to make it verge toward my "precise" pole of the
spectrum. Why has this aspect of Darwin's thinking been excluded? Because the theory doesn't address
these aspects of the phenomenon. There is nothing wrong with that - no theory can do everything. But
the fact that speciation is excluded does support Edser's position that definitions of phenomena may
be contaminated by the theory.
>
> What are the possible causes of these changes that would form part of evolutionary theory? One of
> them is a Lamarckian form of evolution where the heritable characteristics of a population changed
> within an individual in response to the environment. In theory, we could imagine that there would
> be no differential survival of organisms in this this type of evolution (i.e. every individual
> gave rise to one and only one descendant). The heritable characteristics of the population would
> change over time and the population would evolve.
>
> We know that this isn't the way that evolution occurs but it isn't ruled out by the definition of
> evolution. That's the way it should be.
>
> What are some other possible causes of changes in heritable characteristics? One of them is
> mutation. New alleles (mutations) arise in a population at measurable frequency. This is an
> important part of evolutionary theory and it is consistent with the definition of evolution since
> the definition doesn't specify causes. Your definition does specify a cause. It says ...
>
> "Evolution is change in gene frequencies caused by differential survival and reproduction of
> organisms."
>
> Is it your intention to eliminate mutation and possible Lamarckian explanations by defining them
> out of existence? Why do you want to do this?
>
I certainly wouldn't want to rule them out in a loose definition of the phenomenon. But if my theory
ruled them out, or if my theory just didn't cover these possibilities, I might word the precise
definition as I did. You did ask me to provide a precise definition, remember?
Incidentally, your preferred definition suffers from a similar problem. The phrase "spread over many
generations" seems to define saltationist explanations out of existence. I suspect this wasn't your
intention. But, nonetheless, your definition seems to be something of a creature of a theory that
DOES rule out saltation. Also, did you intend to rule out gene changes within a generation - such as
Lamarckism? Or, does the "spread over many generations" simply indicate the delta-t that should be
used in measuring change? There are probably changes in gene frequency that take place over the
course of a year in a population of annual plants. That certainly isn't "evolution", but I'm not
sure it is a good idea to exclude explanation of this phenomenon in the definition.
These are quibbles, to be sure. My intention is not to attack your definition or to attack you. I'm
simply pointing out that it can be difficult to come up with definitions that are totally
independent of theory.
> I realize that we're quibbling over details. Perhaps it would be best to think of your proposed
> definition as a simple variant of what most of us agree is the standard definition. I really don't
> see that it's profoundly different in spite of the fact that I think it's a bit too restrictive.
>
> > Since Kimura and Hamilton, we have better precise definitions that work even better. That is
> > because we have better theories. Of course, if you prefer to think of precise theories as not
> > containing any reference to causation, then Kimura and Hamilton don't change Fisher's precise
> > definition, in your view. But you have to admit that the definition of evolution HAS changed
> > since Darwin. And, for that matter, the definition of "gene" has changed since Fisher.
>
> I don't see that Kimura and Hamilton had any effect on the definition of evolution.
No effect on your definition. Kimura, at least, has an effect on the precise, theory-laden
definition that you asked me to provide. My precise definition should probably be changed to
accomodate Kimura.
> I also don't see any significant different between the definition that I prefer and what Darwin
> thought of as the minimal definition of evolution. Darwin didn't know as much as we know about
> genes and how they work but he did know about heritable characteristics and why they were
> important in evolution. In fact, he proposed a theory of genes.
>
> >> Please explain why your "precise" definitions isn't mentioned in the textbooks on evolutionary
> >> biology.
> >
> > Clearly, it IS mentioned in textbooks.
>
> Okay. Let's assume that your proposed definition is mentioned in textbooks. Why is it more
> "precise" than the one I prefer? Is it simply because you use the word "genes" instead of
> "alleles" or "heritable"? Or, is it because you specify a cause, a cause which seems pretty
> nebulous to me.
Because it specifies a cause. That is by the definition of "precise" which started this discussion.
A hope I have made clear that my "precise" definition is an obsolete one (since Kimura) and that
"precise" is not better than "loose" in my mind.
>
> >> I'm sure you must have such a "precise" definition in mind because if you don't your posting
> >> doesn't make sense. It would be a great help to me if someone could come up with an example
> >> definition instead of making vague accusations against the standard one. (I'm expecting a
> >> comparision between your precise definition of evolution and the "loose" one that evolutionary
> >> biologists use. Which one is better, in your opinion, and why?)
> >
> > The precise definition is best for explaining, testing, and understanding the theory. The loose
> > definition is best for expanding, improving, and overthrowing the theory.
>
> Evolutionary theory is supposed to explain the causes of evolution. In order to do this we have to
> have a common understanding of what evolution is. That's the purpose of a definition. There may be
> many possible causes of evolution. Some of them will be right and some of them will be wrong. It
> seems to me that you are advocating that we change the definition of evolution according to the
> best available evolutionary theory of the day. If we had done this in the past then the definition
> of evolution might include a cause such as natural selection and that would be extremely self-
> serving and wrong. It would restrict evolution to only those causes that most people accepted in
> any given year and eliminate by fiat all competing causal theories.
I agree with this. That is why we need loose definitions. Precise definitions, contaminated by
theory, shouldn't be used to delimit the discipline.
>
> There are some people who would like to do this. They would like to "define" evolution as changes
> caused by natural selection and eliminate random genetic drift as completely irrelevant to
> evolution. Some of these same people want to rule out neutral changes. They don't think they
> represent real evolution. In my opinion this is a serious error.
Are you sure that "some people" are claiming that neutral changes don't happen? Or are they simply
excluding them from the theory because the theory doesn't cover them? (Much like the way you have
excluded speciation from your definition.) I can't say that I have sufficiently come to understand
"some people"s thinking yet to answer my own question here.
Obviously, it is an error to claim that neutral changes don't happen. It is also an error to claim
that they are irrelevant, unless you specify clearly what you think they are irrelevant to. And
finally, it is reckless to assume that they won't contaminate the testing of a narrow theory that
excludes them without doing a careful analysis. But, having done that analysis, you have just
constructed a testable theory that includes neutral changes. So, in the final analysis, I guess I
would have to agree with you that "some people" are wrong.
> > Personally, I am disappointed that the modern precise definitions of evolution don't explicitly
> > mention speciation.
>
> This is a problem. If we want to cover all levels of evolution then we should also deal with the
> birth and death of populations. That's why I emphasize that the definition is a "minimal"
> definition designed to distinguish between processes that resemble evolution at the population
> level and process that really count as evolution.
>
> >> > I encourage doubters to produce a list of phenomena and check this conceptual framework
> >> > against their list. My list included the origin of life, the origin of the moon, human
> >> > monogamy, photosynthesis, and others. In each case, I notice that the competing theories
> >> > (that I am aware of) agree on a loose definition of the phenomenon, but introduce more or
> >> > less subtle differences in their precise definitions. Particularly interesting to Einstein
> >> > fans is one of Larry's examples - gravity.
> >>
> >> I don't understand these sentences. Could you elaborate? Are you referring to *definitions* of
> >> "the origin of life" etc. I would *define* "the origin of life" as "the point in time when the
> >> first life began." Do you have something else in mind?
> >
> > I would change your loose definition to "the process by which the first life began" for
> > starters. The process is what people are interested in. But then someone is going to ask for a
> > definition of "life". And people like Francis Crick are going to foul things up by offering
> > theories like "Directed Panspermia". Should we ask for the origin of life on Earth? People like
> > Cairns-Smith are going to come up with theories of organisms built from clay that are ancestral
> > to modern organisms only with some stretching of concepts. Every theory forces some revamping of
> > any attempted precise definition.
>
> Please offer a definition that illustrates the problem and how you would solve it.
Are you asking me for a precise definition of the origin of life? I can't provide one, because I
claim that a precise definition is a creature of the theory, and I don't have a theory of the
origin of life.
If Crick believed that his Directed Panspermia solved the problem, then he might have been tempted
to provide a precise definition of the origin as "the process whereby life first appeared on earth".
His causal explanation: aliens did it.
As a loose definition of the origin of life, I would prefer "the process by which autopoeisis came
into existence". Other people would prefer "the process by which reproduction with heritable
variation came into existence". Cairns-Smith has a theory which satisfies the second definition. I
believe that the theory must be extended to satisfy the first definition, to be taken seriously.
Does a theory of the origin of life have to explain protein synthesis and the code?
My point is that different theorists will prefer different definitions of the phenomenon.
>
> > You really never understand exactly what it is you are trying to explain until you explain it!
> > Thereby, you produce a precise definition. But even before you succeed, you had a goal in mind
> > for what you were going to explain - that vague goal is the loose definition.
>
> Hmmmm ... I really don't see your point. The distinction between what you call a "loose"
> definition of evolution and a "precise" definition of evolution is very close to nitpicking.
> Nitpicking can be fun but there's a risk that we miss the main point.
For the phenomenon of evolution, there is some truth to what you say. Evolution is a well studied
phenomenon; we have some concensus on how narrow the definition should be; there are clearly
multiple causes to be covered; and we may be able to come up with a single definition (like yours)
that doesn't mention theory, but provides exactly the right amount of narrow focus.
If you and John had restricted your controversy to the definition of evolution, then maybe I
wouldn't have been tempted to comment. But towards the end, it seemed that you two were talking
about the relationship between theory and definition in general. And, for a lot of phenomena, I
think the relationship is pretty muddy.
>
> >> What about "gravity"? The word is normally *defined* as something like "the attractive force
> >> between massive bodies that is proportional to the product of their masses and inversely
> >> proportional to the square of their separation." This doesn't look to me like a definition
> >> that's a creature of theory. What did you mean?
> >
> > Actually you are giving Newton's precise definition of *gravitation*, not of gravity. Gravity is
> > the force here on earth that makes apples fall and it was known long before Newton.
>
> I think this discussion is getting way too semantic for my liking.
Gee, I'm sorry. Semantics has no place in a discussion of the meaning of the word "definition". ;-)
I hope you didn't interpret my distinction between gravity and gravitation as some kind of putdown.
I seized on it only because the fact that gravity can be (mostly) reductionistically explained by
gravitation seemed like it would help me make my point.
>
> > Galileo provided the first quantitative theory of gravity. Newton reduced gravity to a special
> > case of a new theory - gravitation. Einstein showed that Newton's gravitation, along with the
> > laws of conservation of momentum and energy, are only approximations to a more general theory -
> > General Relativity. The loose definition of gravity as a tendency of things to seek the center
> > of the earth is still valid, but the precise definition changes with each new theory. Even if
> > you leave out Einstein, gravity now includes a mix of the centripetal force due to gravitation,
> > the centrifugal force due to the earth's rotation, and small changes in the shape of the earth
> > caused by lunar and solar tides.
> >
> > Incidentally, Newton had a loose definition of gravitation before he had a precise one. He knew
> > there was an attractive force from Copernicus, but it took a careful analysis of Kepler's laws
> > to determine that it must be an inverse square force.
>
> Please give me your preferred *definition* of gravity so I can try and see what you're talking
> about. It seems to me that you are confused about the difference between a definition of gravity
> and the cause (explanation) of gravity.
I gave an ancient loose definition above - a tendency of things to seek the center of the earth. A
good precise definition for Galileo's time might have been - the acceleration of 9.8 m/sec^2 toward
the center of the earth experienced by an unsupported object. At the time, there were no good causal
explanations. However, I think that it is clear that Galileo had created a THEORY of gravity simply
by giving a precise definition of the phenomenon. Not all theory is reductionism.
Newton, though, did provide a successful reductionist (causal) explanation, and also redefined
the phenomenon - pointing out that gravity varies with altitude and with latitude. New theory -
new precise definition of the phenomenon. Please don't ask me to put Newton's precise definition
into words.
Newton's law of gravitation has a precise definition much like the one you offered above. Did it
include a causal structure? Newton didn't think so. "Hypotheses non fingo". Later, Einstein provided
one, but only by changing the precise definition to include gravitational effects of rotation,
gravitational effects of massless objects, and other refinements.
My point is simply that precise definitions of phenomena change with the theories that support them.
I might ask, by the way, which came first - your preferred definition of evolution, or your
preferred theory of evolution? You have crafted your definition to accomodate future changes in
theory, but I suspect that careful scientists of a couple of generations ago, knowing less than you
know of modern theory, might not have come up with the definition that you now propose.
My final point, which I'll hope you agree with, is that rigid epistemological methodologies don't
work. Saying that THE definition MUST be theory-neutral is as much of a mistake as saying that THE
definition MUST be part of the theory. The progress of science is much messier than that.
No, let me correct that. That was my ORIGINAL point, and my only point. I hope we can let this
subject die. You are welcome to have the last word, if you want.
Jim