|
Author
|
Topic: Speciation
|
warren_bergerson
Member
Member # 262
|
posted 16. April 2004 09:44
I am having a hard time reconciling the conventional genetic determinism concept- phenotype classes (species) are caused by permanent and universal causal processes operating on genotype classes- and the observed existence on distinct species.
As John Davison has noted, there does not appear to be any convincing evidence that sexual reproduction is currently producing new species. This observation raises the question of how and why species (apparently) arose in the past. But of even more immediate interest, IMO, it is the question of how and why distinct species can exist in the world today. If the development of organisms from sets of genes is controlled by natural laws which existed from the first occurrence of DNA, why do there exist species boundaries which prevent successful interbreeding?
Dr. Davison is suggesting, if I understand him correctly, that the absence of current observable sexual reproduction speciation events is evidence supporting his proposed hypothesis. However, it is not clear to me how the lack of observable speciation contradicts any neo-Darwinian theory which accepts the conventional genetic determinism assumption.
It should be noted it appears possible to explain speciation if you replace the conventional ‘permanent causal process’ genetic determinism assumption with a teleological causal paradigm genetic determinism assumption. Using the teleological assumption, speciation would be predicted to occur when speciation benefits the goal of survival. As a simple example, speciation might be beneficial if there existed two food sources which required incompatible physical traits (one type of food was best exploited by large individual and the second was best exploited by small individuals). Under such conditions it might be beneficial to avoid interbreeding which would be likely to produce intermediate sized individuals not adapted to either food source.
If evolution is an intelligent goal directed process, and if genotype to phenotype transformations are modifiable causal paradigms, then, I suggest, speciation might be produced by creating incompatibilities in the developmental processes or mechanisms of the two sub-groups. In other words, if developmental processes are modifiable or programmable, then it should be possible to create two developmental processes which are similar but incompatible.
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 16. April 2004 15:57
Let’s try this from a slightly different angle. First, it is very easy to do a pencil and paper design of a process or mechanisms which produces speciation. All you need is a process or mechanism with the logical form “If gene or allele X is present (or better yet if X is absent) then produce effect Y which prevents development of organism capable of reproducing”. Since we know that different species exist, we know that such processes or mechanisms must logically exist.
Since we know or can reasonably assume that developmental processes can be modified, we can reasonably conclude that developmental processes can be, and have been modified to produce speciation. If you prefer, this logic can be expressed as a prediction of the form ‘life forms have the capacity to design and implement species boundaries’.
All that is required for this type of explanation of speciation is accepting the perspective that the developmental processes by which genotypes determine phenotypes are best viewed as programmable causal paradigms rather than permanent causal processes.
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 17. April 2004 08:44
Just as a reminder, RM&NS type Darwinian theory requires the existence of 1)a deterministic causal relationship from genotype as cause to phenotype as effect and 2)a deterministic causal relationship between phenotype as cause and genotype as effect. The first causal relationship asserts that changes in genotypes called mutations cause changes in phenotype. The second causal relationship asserts that changes in phenotype produced by natural selection cause changes in genotype. The existence of these two types of causal relationships is established in traditional evolutionary biology as an assumption called the genetic determinism assumption.
The evidence, including the evidence relating to speciation, would suggest that the causal relationships between genotype and phenotype are dynamic, modifiable or programmable. In my terminology, genotype to phenotype causation involves a dynamic causal paradigm rather than a static causal process.
The existence of dynamic genetic determinism is not incompatible with the development of predictive scientific theories of evolution. All that is required is that you define the processes responsible for producing changes in the genotype to phenotype mapping. To my knowledge, no one has proposed a hypothesis addressing the issue. No current hypothesis of evolutionary or genetic change explicitly recognized that genetic determinism is dynamic and teleological. There is, therefore, no current hypothesis of evolutionary or genetic change which can fit the available evidence.
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 17. April 2004 11:37
For the benefit of lurkers, let me suggest that warrenbergerson's unawareness of the literature of biology leads him to make some strange remarks about it.
warrenbergerson wrote quote:
Just as a reminder, RM&NS type Darwinian theory requires the existence of 1)a deterministic causal relationship from genotype as cause to phenotype as effect and 2)a deterministic causal relationship between phenotype as cause and genotype as effect. The first causal relationship asserts that changes in genotypes called mutations cause changes in phenotype. The second causal relationship asserts that changes in phenotype produced by natural selection cause changes in genotype. The existence of these two types of causal relationships is established in traditional evolutionary biology as an assumption called the genetic determinism assumption.
I suggest that warrenbergerson do a brief Google search on "genetic determinism assumption." And I'd like to see a reference or two for #2. Natural selection is generally held to produce changes in the frequency of alleles in the population, not changes in the genotypes of individuals. Is that what warrenbergerson means?
warrenbergerson further wrote quote:
The existence of dynamic genetic determinism is not incompatible with the development of predictive scientific theories of evolution. All that is required is that you define the processes responsible for producing changes in the genotype to phenotype mapping. To my knowledge, no one has proposed a hypothesis addressing the issue.
Sure they have. A slew of such hypotheses are embodied in a whole subdiscipline called developmental evolutionary biology (sometimes "evo-devo"). See here for a brief description, and here for a longer (albeit already dated) introduction.
RBH
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 17. April 2004 17:14
I am glad to hear that RBH agrees that biologists recognize the fact that the genetic determinism relationship is dynamic or programmable.
While I am aware that biologists are willing to recognize that developmental processes are dynamic, modifiable, or programmable, I was not aware that biologists recognized that genetic determinism was dynamic. I was not, in fact, aware any of the conventional sciences recognized anything other than permanent and universal determinism or stochastic determinism.
In some scientific circles, I believe, it would be considered logically inconsistent to claim both 1)genotype determines phenotype and 2)the relationship between genotype and phenotype can be changed or reprogrammed.
My proposed form of teleological determinism does allow for relationships which are considered deterministic and which have the appearance of being programmable. I was not aware that this form of determinism was recognized by any of the conventional sciences. Also, teleological determinism involves recognition of both the concepts of purposeful causation and intelligence. I would be surprised if most biologists would be willing to accept that genetic determinism is the result of purposeful intelligence.
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 17. April 2004 17:56
warrenbergerson wrote quote:
I would be surprised if most biologists would be willing to accept that genetic determinism is the result of purposeful intelligence
So would I be surprised. Research on the naturalistic evolution of that programming is in progress in a number of labs.
warrenbergerson further wrote quote:
In some scientific circles, I believe, it would be considered logically inconsistent to claim both 1)genotype determines phenotype and 2)the relationship between genotype and phenotype can be changed or reprogrammed.
Actually, the whole point of evo-devo is an explication of the denial of #1 sensu strictu and the elaboration of how #2 occurs. No one I know holds a strict 'linear' genetic deterministic view like that suggested by warrenbergerson as #1. It has been known for some years that there is a many-one relationship between genotypes and phenotypes. There's a growing body of research on the evolution ("programming") of the mapping - evo-devo again.
RBH
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 17. April 2004 21:34
It's many-to-many. Identical genotypes can give rise to different phenotypes if conditions are different (e.g. identical twins). Different genotypes can give rise to essentially identical phenotypes (e.g. male C. elegans with and without egl-1, a gene involved in the development of the egg-laying system, or anything involving a recessive allele, etc.).
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 18. April 2004 01:11
Oops. Right. Many to many it is. I'm going to cease posting until I can quit taking the painkillers.
RBH
[ 18. April 2004, 01:13: Message edited by: RBH ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 18. April 2004 10:41
It seems doubtful if many or many type relationships would qualify as deterministic. At issue here is the difference between dynamic relationships and deterministic relationships. Dynamic or programmable relationships have the appearance of being indeterminant or the opposite determinant.
For the sake of lurkers who may not be familiar with the concepts, a relationship is characterized as determinant if it can predict future events. A genotype to phenotype relationship would be deterministic if we can predict future occurrences of phenotypes based on genotype. If, as apparently recognized by developmental biologists, the developmental processes are dynamic or programmable, then it may not be possible to predict future phenotypes from a specific genotype because the developmental process may have been modified or reprogrammed.
We have neo-Darwinian ‘theory’ which ‘assumes’ the genotype to phenotype relationships are deterministic, and we have observations from developmental biologists suggesting the genotype to phenotype relationship is dynamic or programmable. This is what we might call an unresolved contradiction or an ‘apparent’ unresolved contradiction. To some people, such contradictions would be resolved by accepting the observed facts and by rejecting the theory.
The dynamic versus deterministic genotype to phenotype is an ‘apparent’ contradiction because there may exist some resolution. For example, the apparent contradiction could be resolved if 1) you could successfully predict the dynamic reprogramming which occurs and 2) you could identify the processes responsible for producing the changes. My teleological determinism proposal asserts 1) you can predict the changes that will occur based on purpose or the goal of survival, and 2) you can explain the processes responsible for the purposeful changes by a form of dispersed processing which is called intelligence.
The issue of speciation provides an simple illustration of the concepts involved. Teleological determinism would predict that speciation will occur if speciation is beneficial to the goal of survival. As discussed earlier, speciation might be beneficial if it allowed a single group to exploit two food sources requiring incompatible physical traits. If speciation is beneficial, then intelligent processing should be capable of designing and implementing a reproductive boundary which creates a species. In effect, I am predicting that speciation will result from purposeful intelligent design of reproductive boundaries.
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 18. April 2004 11:28
The movement of billiard balls on a table is deterministic, but I have never seen two identical breaks. It is in this fashion that one genotype must lead to multiple phenotypes in many cases: the genotype encodes a system that is affected by the world, and the world is not constant. Whether or not the process is truly stochastic (i.e. whether or not quantum fluctuations during development affect the final phenotype), it is effectively stochastic.
One need not invent a new category of "dynamic relationships" that is separate from "deterministic relationships". However, the mode of resolution is thus far turning out to be less exotic than warren_bergerson speculates. For instance, while alternative development paths are typically adaptive, one can predict them on the basis of external inputs and signaling cascades more readily than by an ill-defined "goal of survival", and the processes involved are molecular signaling cascades which one could define as intelligence, but doing so would not be very instructive. These cascades, insofar as we have been able to determine, are just chemistry and operate deterministically/stochastically.
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 18. April 2004 12:54
There is an apparent logical inconsistency between the genetic determinism assumption used in neo-Darwinian RM&NS type theories and the observation by developmental biologists that developmental processes are dynamic or modifiable.
It might be useful to note that this apparent logical inconsistency is potentially a problem for neo-Darwinian theory. It is not necessarily a problem for the study of developmental processes. As happens in practice, a biologist can study a particular developmental processes or mechanisms as it exists in a specific species at a specific point in time. The scientist may recognize that the process being studied is dynamic, modifiable, or programmable but he need not directly address the issue as long as the process is static or deterministic within the context of the study being performed. The fact that the scientist studying developmental processes can deal with dynamic genotype to phenotype transformation does not mean this dynamics is not a problem for theories based on the genetic determinism assumption.
As I discussed earlier, the causal paradigm concept does not deny the existence of physical causal pathways. Instead, the causal paradigm concept suggests that there are many different such pathways, and that it is more useful and practical to develop predictive models and hypotheses in terms of the cause and effect relationship rather than in terms of specific causal pathways.
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 19. April 2004 03:24
It's just a computational problem, not an intrinsic contradiction. Most models don't include such effects because they're a pain in the neck to compute and we don't know enough to bound them properly.
I'd like to see more studies done where "indeterminism" is explicitly included. But it's not the kind of thing that brings an entire theory crashing to its knees.
Calling something a "causal paradigm" doesn't make the computational difficulties go away.
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 19. April 2004 07:55
As stated previously, recognition that the genotype to phenotype causal pathways are modifiable or programmable defines an apparent logical contradiction with the genetic determinism assumption on which neo-Darwin theory depends.
Theories, as Rex notes, do not collapse just because they contain or appear to contain a logical inconsistency. If order for an apparent contradiction to result in the collapse of a theory, the scientists ‘deciding’ on the acceptability of the theory must 1) recognize the existence of the apparent contradiction, 2) must accept that there is no resolution of the apparent contradiction, and 3) must accept that the existence of a logical contradiction is a valid basis for rejecting a theory. Also, it appears that in many instances scientists are reluctant to reject one theory unless someone presents and acceptable alternative. It would appear based on the discussions here and elsewhere that most biologists are not convinced that dynamic genotype to phenotype pathways are a sufficient reason for rejecting neo-Darwinian theory.
It is well known from computer programming that you can build multi-layer or hierarchy programs, where the highest level program is a static or permanent program which controls the dynamic changes or apparent dynamic changes in lower level programs. Whether or not it would be possible or practical to build a high level static program, model or hypothesis which is capable of controlling, explaining or modeling the dynamic changes in developmental pathways is another interesting question. If it is possible or practical to develop such a controlling model or algorithm, such an algorithm or program will almost certainly be far more complex that the simplistic RM&NS mechanism.
I am proposing that there is a type or class of high level static program which can control and predict changes in developmental pathways. Such a static program I suggest is a system of purposeful intelligence. An abstract mathematical model of such a system with purposeful intelligence is what I call a LEPS system.
IP: Logged
|
|
nosivad
Member
Member # 767
|
posted 19. April 2004 07:56
Warren Bergerson's opening statement clearly defines the problem. Why are there species boundaries? If natural selection were now or in the past a factor in speciation there could be no taxonomy. Everything would be a morphological continuum. Are we to believe that gradual transformations were the rule in the past when everything we see now indicates the opposite? I think not. Since macroevolution is no longer evident, we have a responsibility to postulate mechanisms which might have produced what we now observe. That is exactly what the semi-meiotic hypothesis may be able to explain. The Darwinians have rejected the hypothesis out of hand without testing it. Let me try to explain why. Not only do they not consider that there might be endogenous and prescribed laws which have governed evolutionary change, they deny that such laws could possibly exist, or more exactly, could have existed. If the restructuring of existing genetic information could result in the production of a new species, the entire Darwinian model would collapse. Actually, as I explained on Fernando's thread, speciation (in reverse) has already been demonstrated through chromosome restructuring in yeast. That finding alone should sound the death knell for neoDarwinism.
I happen to believe, with Godfrey Hardy, that mathematics exists independently of the human condition. I propose that the same is true for all of chemistry, physics and biology including ontogeny and phylogeny. Just as mathematics, chemistry and physics all have operational rules and restraints, so must all of biological phenomena have the same. The major difference is that the rules which have operated to produce organic change have not yet been clarified. One thing is for certain - chance has not played a significant role in either speciation or the production of any of the higher categories.
This in no way is in conflict with genetic determinism. I take exception with Rex Kerr's interpretation of the significnce of studies on separated identical twins. The most remarkable feature revealed by such studies is the extent to which they support a hard-wired role for heredity in determining not only our physical features but the way in which we view the world on matters such as left versus right politics, abortion, the death penalty, a belief or disbelief in a creator and attitudes toward authority generally. William Wright's book "Born That Way" should be given serious consideration by anyone interested in the nature/nurture issue. Thus, just as all of evolution may have been predetermined in accord with laws, as Leo Berg first suggested, so also may be the way we interpret reality. In short, there may very well be a genetic basis for ideology. That can go a long way toward explaining the polarization that continues to interfere with progess in understanding the great mystery of evolution. We may be, as Wright's title suggests, born that way. In any event there is no reason to discard genetic determinism. It cannot be denied without abandoning everything we have discovered in the last hundred years. I, for one, am certainly not prepared to take that step.
[ 19. April 2004, 10:09: Message edited by: nosivad ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 19. April 2004 16:27
Genetic determinism is generally considered an assumption. At least in some circles, assumptions are required to be consistent with relevant facts.
At least in some applications, it is acceptable to recognize an assumption as an approximation of the true relationship. In some type of applications with which I am familiar, it was considered appropriate to perform sensitivity testing to demonstrate that an approximate assumption was not distorting the results being reported.
It seems reasonable to conclude that in at least some applications, even simple one to one mapping forms of the genetic determinism assumption provide a reasonable and appropriate approximation.
There are other applications where, at least IMO, the it is not clear that the use of the traditional genetic determinism assumption is justified. As discussed, scientists studying developmental processes generally recognize that genotype to phenotype transformation processes are modifiable or controllable. It is also generally recognized that genotype to phenotype transformation processes depend on environmental conditions. Technically, this can be interpreted to mean the phenotype is determined by genotypes plus other factors.
With respect to the specific issue of speciation, if you 1)abandon the genetic determinism assumption and 2)recognize that genotype to phenotype transformations are modifiable, then 3)you can design relatively simple mechanisms which could produce speciation( mechanisms which prevent the development of offspring capable of reproducing when members of two sub-groups interbreed). If scientists can identify the existence of such speciation mechanisms in nature, and if scientists can design situations where these mechanisms arise and are implemented in nature, then you have a model or explanation of how speciation can occur in organisms which reproduce sexually.
IP: Logged
|
|
|
Topic Closed
Printer-friendly view of this topic
