|
Author
|
Topic: Dynamic Best Practices Topology
|
warren_bergerson
Member
Member # 262
|
posted 28. April 2003 12:39
The term topology can be used to refer to the mathematical/logical structure used to express or frame a related set of scientific models and theories. The ‘dynamic best practices topology’ is proposed here as an appropriate topology for the scientific analysis of life forms. This includes the analysis of the processes by which life forms create intelligent designs.
The dynamic best practices topology starts with a standard space-time structure. Within this space-time structure are defined entities or systems with the ability to perform a type of information processing labeled biological information processing (accept or record input from the environment external to the entity, process the information as defined by a processing algorithm, and generate output into the external environment). The information processing capacity of the defined systems is represented by programmable logic machines.
The mathematical space time universe with entities or systems performing biological information processing is defined as having the following properties:
1. MULTIPLICITY- Processing algorithms can take multiple forms- The set of possible algorithms for entity x at time t is denoted by Fxt.
2. DYNAMIC MULTIPLICITY- The set of possible algorithms can be different for different entities and for the same entity at different times.
3. FUNCTIONALITY- Processing algorithms perform tasks, functions or purposes.
4. EFFECTIVENESS- Some of the possible algorithms are more likely to successfully perform task or function than others. - The form most likely to perform a task, function, or purpose at a point in time is called best practices form or algorithm.
5. DYNAMIC BEST PRACTICES- The best practices form is dynamic. The algorithm most likely to perform a task can be different at times and in different entities due to either a)differences in environmental conditions or b)differences in the set Fxt of possible algorithms. For an entity of system, the set of best practice algorithms over a period of time is called dynamic best practices set.
6. SURVIVABILITY- The effectiveness of an entities processing algorithm(s) effects the likelihood an entity or system will survive.
7. COMPETITIVE ENVIRONMENT- The survivability of one entity or group of entities impacts the survivability of other entities or groups of entities.
The above properties are relatively standard features used in modeling systems which perform information processing. Note that these properties are being defined as properties of the mathematical structure or topology rather than properties of the real world. This is believed to be an important distinction. It is often pointed out that there is no evidence of ‘purpose’ in the universe. However, it is equally true there is no evidence of causation in the universe. Causation and purpose, as viewed here, are properties of a mathematical structure or topology which is defined in order to construct scientific models and theories. While the existence or non-existence of causation and purpose in the real world can be debated, there seems no reason to doubt that scientists and mathematicians can impose these properties on the topology of an abstract mathematical universe.
While the topology defined by the above properties represent a relatively standard approach to modeling biological information processing, it is not clear that testable, predictive models and theories can be formulated using such a topology. In order to create predictive theories relating to life forms, it appears that it is also necessary to define the following properties or principles:
THE EXPECTED ALGORITHM PRINCIPLE:
In the ideal form of the mathematical universe being defined, a universe without noise or inefficiency, given 1)a task or function to be performed, 2) multiple methods of processing information to perform the task or function, 3)a dynamic environment which produces changes in task requirements, and 4)dynamic technology which can change the methods available to perform the task or function, then biological information processing will follow dynamic best practices.
THE STABILITY PRINCIPLE:
In the competitive form of the mathematical universe being defined, any entity, system or group not performing biological information processing based on dynamic best practices is likely to be replaced by an entity, system or group performing biological information processing based on dynamic best practices.
These principles or assumptions ‘define’ dynamic best practices as the expected or efficient form of biological information processing. Furthermore, any form of information processing other than best practices is defined as unstable. It is again worth noting that the best practices assumptions, principles or definitions are defined or imposed in order to make it possible to construct predictive models and theories in situations where the rules or ‘laws’ governing information processing are dynamic, and where there is no absolute basis exists for predicting what changes will occur.
The best practices topology is interesting, I suggest, because:
1. It makes it possible to formulate predictive models and theories for dynamic information processing. Such predictive theories are not logically possible using the traditional ‘permanent and universal’ causal topology.
2. Darwinian and neo-Darwinian theories are not compatible with the best practices topology. Evolutionary change is a form of biological information processing. Darwinian theories suggest a fixed process. The dynamic best practices paradigm suggest a continually changing process.
3. Scientific analysis is a form of biological information processing and can be analyzed assuming the best practices topology. If as seems likely current academic science is not a best practices set of procedures, then current academic science is likely to be unstable and replaceable. [The best practices topology provides a basis for evaluating current scientific procedures.]
SUMMARY
I have defined or at least outlined the key features of a logical structure or topology which could, I suggest, be used to formulate sets of related scientific models and theories dealing with life forms. At the very least, the dynamic best practices topology provides some interesting contrasts to current academic science.
IP: Logged
|
|
Evan
Member
Member # 164
|
posted 28. April 2003 18:52
Warren writes,
quote:
However, it is equally true there is no evidence of causation in the universe.
I'm curious as to what Warren means by this? Finding cause-and-effect relationships is one of the things science is all about. Of course, there is a limit to our ability to trace causes back - the traditional difference between proximate and ultimate causes comes into play.
But Warren, what do you mean when you say there is no evidence of causation?
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 28. April 2003 19:01
What is a "best practice" as opposed to a non-best practice? Is it possible to give a definition of the form, "X is a best practice if and only if (list of conditions that need to be met)"?
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 29. April 2003 07:44
Evan,
We can observe ‘B follows A’, we can observe "b a member of B always follows ‘a’ a member of A", we can successfully predict that "b a member of B will always follow ‘a’ a member of A". We can not, however, observe ‘A causes B’. ‘A causes B’ is a logical or mathematical concept, or interpretation imposed on real world observations by the observer". I am using the term topology to refer to the set of logical/mathematical concepts or structures used or imposed in performing a particular type of scientific analysis.
The point of my comment about causation, is that if it is legitimate for scientists to impose/use concepts like ‘permanent and universal causation’ because the concept is useful, then it is also legitimate to impose/use concepts like dynamic and teleological causation and purpose if such concepts or logical structures are useful.
Rex,
Given 1)a set of functions or processing algorithms Fxt in system x at time t, 2)a task, function, or purpose G, then the best practices algorithm fbp for system x at time t relative to g is the processing algorithm most likely to result in achieving goal G. In an abstract mathematical universe where we can assume perfect knowledge, the concept is readily applied.
In real world modeling, applying the concept of best practices requires defining G, x, t, and F as well as the ability to measure the relative effectiveness of different members of F. The best practices concept is used rather routinely in engineering to compare designs and to compare the operations/effectiveness of different types of machines.
I mentioned briefly yesterday two applications of the best practices concept, one relating to evolutionary theory and one relating to scientific practices.
Traditional evolutionary theory suggests there is some process ‘fd’ which explains evolutionary change for all tasks G, and for all Fxt. Traditional theory suggest the existence of a universal ‘best practice’ evolutionary change process. The dynamic best practices topology suggests that evolutionary processes can continue to change, evolve, and improve, at least with respect to certain tasks. The ‘single universal best practices process’ is thus not compatible with the dynamic best practices topology. In practical terms, a single universal evolutionary process, is not logically compatible with the concept of the evolution of evolutionary processes. Since current evolutionary theory appears to accept both a single universal evolutionary process and the evolution of evolution, it would appear there is a logical contradiction/inconsistency in current theory.
The second application of the best practices concept relates to the procedures of scientific analysis. One of the tasks or functions of scientific procedures is to develop scientific hypotheses and to decide if the hypotheses are sound and valid. The current academic science procedures for performing this task can be denoted by CS (current science). A new competing set of procedures for performing this task could be denoted by NS (new science). If it can be shown that NS is closer to best practices than CS for the task of formulating and evaluating hypothesis, then NS would likely replace CS for that task of formulating and evaluating hypotheses.
Again it is useful to repeat that dynamic best practices is a logical structure or topology imposed on scientific analysis by the analyst because it is expected to produce useful results. You will not find a ‘best practices’ label in nature anymore than you will find mileage markers corresponding to a space time topology.
IP: Logged
|
|
Alix Nenuphar
Member
Member # 686
|
posted 29. April 2003 13:18
Warren:
It would appear that you suggest modeling 'actors' (i.e. intelligent agents) using a three part model: information collection, information processing, and response mechanism.
Could you possibly expand on two points: what functions will you use to model the real-world observational constraints imposed on biological organisms by their structure; and given that you will model information processing via a 'Best Practise' algorithm, can you expand on how such an algorithm will mimic the 'observationally random' choices made by intelligent agents?
[ 29. April 2003, 13:21: Message edited by: Alix Nenuphar ]
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 29. April 2003 14:37
Warren, I agree that under the definition above, Darwinian evolution is exceedingly unlikely to always produce organisms who function with best-practices.
Is there, however, evidence that organisms do always function with best-practices?
(I would note that standard evolutionary theory doesn't reject the idea of the evolution of evolutionary mechanisms, but it is not often included in models given the computational complexity (and uncertainty) that it adds.)
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 29. April 2003 15:37
Rex Kerr wrote quote:
(I would note that standard evolutionary theory doesn't reject the idea of the evolution of evolutionary mechanisms, but it is not often included in models given the computational complexity (and uncertainty) that it adds.)
Which is not to say that it's not a lively area of research, simulation, and theorizing. A Google search on "evolution of evolvability" turns up 1,120 hits. PubMed yields 46, many fewer, of course. Some are very recent, including publications in 2003 concerning the evolvability of varying mutation rates and of genetic linkage patterns. A particularly interesting one for these sorts of discussions in general is titled "Can an arbitrary sequence evolve towards acquiring a biological function?" From the abstract: quote:
This study therefore exemplifies the process of a random polypeptide generating a functional role in rejuvenating the infectivity of a defective bacteriophage when fused to some preexisting protein modules, indicating that an arbitrary sequence can evolve toward acquiring a functional role. Overall, this study could herald the conception of new perspective regarding primordial polypeptides in the field of molecular evolution.
While slightly off on a side track, I found it interesting.
The research on the evolution of evolvability is relevant when assessing warren's assertion that quote:
Darwinian theories suggest a fixed process. The dynamic best practices paradigm suggest a continually changing process.
I'd ask what specifically is it that is fixed in "Darwinian theories" and what is dynamic in warren's "best practices paradigm." It's hard to contrast the two not knowing what properties they're being compared on. warren refers to quote:
"2. DYNAMIC MULTIPLICITY- The set of possible algorithms can be different for different entities and for the same entity at different times.
Is it his argument that because evolutionary theory posits a quite limited set of algorithms, it (the theory) does not account for populations that might display "dynamic best practices"? Or is it that evolution's finite and limited set of algorithms is insufficient to account for observed biological phenomena? I'm not certain what's being argued, nor even what the unit of analysis is - genomes, genome-constructed entities, groups (populations) of entities - that 'has' the algorithms.
RBH
[ 29. April 2003, 15:43: Message edited by: RBH ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 29. April 2003 16:00
Alix,
I am not sure I understand your question, so I apologize if I am addressing the wrong issues. To paraphrase, I am suggesting modeling the behavior of biological entities using information processing. I would characterize all such entities as intelligent agents.
There are all sorts of constraints applicable to the input used in biological information processing. How one would address these constraints would depend on the particular situation being addressed.
Consider as an example the comparison of two competing forms of scientific analysis, CS (current academic science) and some as yet undefined alternative form of scientific analysis NS. NS and CS are to be compared on their ability to formulate and ‘correctly’ evaluate/validate scientific hypotheses.
Since in science information is required to be publicly available, both CS and NS would be expected to operate on the same input. Input constraints would impact both processes equally. The differences between CS and NS would be the criteria, methods and standards used in constructing and evaluating hypotheses.
Quote: given that you will model information processing via a 'Best Practise' algorithm, can you expand on how such an algorithm will mimic the 'observationally random' choices made by intelligent agents?
Does this refer to free will?. For a given set of input and a given set of output choices, different entities will make different decisions or choices. This phenomena occurs even when a best practices algorithm operating on the available information would select a single option.
The dynamic best practices topology would interpret or interpret such a result in terms of noise or inefficiency. The decision produced by the ‘best practices algorithm’ would be the expected or predicted result and the variance from the expected result would be interpreted as noise or inefficiency. As would be expected, the level of noise or inefficiency generally decreases as the importance or value of the decision increases.
Note the in the topology defined best practices algorithms are dynamic or changeable. A system with no variance or deviation from best practices would have a limited ability to adapt to changing conditions. The complete lack of deviation or variance from ‘current best practices’ would probably not be a ‘dynamic best practices algorithm’.
As a general rule, biological information processing appears to be highly efficient. The deviations from dynamic best practices do not appear to be substantial even when best practices algorithms are changing rapidly. (In neurons, as an example, the ‘best practice algorithms’ can change in time frames measured in milli-seconds or less. The apparent large degree of noise, inefficiency or free will in human decision making is, I suggest, due to identifiable factors other than noise or inefficiency.
Rex,
To clarify a couple of points. The dynamic best practices topology asserts, assumes or defines that ‘under ideal conditions( in the absence of noise or inefficiency) biological information processing follows dynamic best practices’. Predictive scientific hypotheses by convention predict not what will actually happen or be observed, but what is expected to happen under idealized conditions. Scientific theories developed under the proposed topology would ‘predict’ the ‘dynamic best practices result’. Since there is always noise and/or inefficiency in any real world system, actual observed results can and do depart from dynamic best practices.
The dynamic best practices principle is a solution to the problem of making reliable predictions for dynamic or changeable processes. Based solely on observations of past processing, you can not logically predict future output if the algorithm responsible for processing is dynamic or changeable or programmable. However, if you know the goal or purpose of the information processing, the available input, the available output and the response options, the available or possible processing algorithms and you assume processing will follow best practices, then you can predict the output that would be produced by best practices.
Analysis using dynamic best practices topology suggests inconsistencies in Darwinian theory. First, it is logically impossible, at least in the dynamic best practices topology, to have 1)a precisely defined permanent and universal evolutionary process which explains /predicts all evolutionary change and 2)a dynamic progressive evolutionary process that is continually changing and improving. With apologies to Dr. Pangloss, "This is the best of all possible worlds and it is getting better every day", is not a logically consistent position.
On a more practical level, we come to questions like "If evolutionary processes evolve, what are the limits to the improvement that can be achieved?". Can, for example, evolutionary process evolve types of selection other than natural selection? Can evolutionary processes evolve non-random mutation processes? Can evolution evolve evolutionary process which operate to produce adaptive change within the lifetime of the organism? Analysis and theory construction using the proposed topology, I suggest, produces predictive hypotheses which address these issues. Hypotheses which, again I suggest, which are not compatible with Darwinian theories.
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 29. April 2003 21:48
I grant that actual observations show deviations from best-practices algorithms, and that one always expects some deviation from theory under ideal conditions and actual results.
However, what is missing is a reasonable indication that evolution should, ideally or otherwise, follow best-practices. For instance, if you're not already perfect, having a fairly static evolutionary process that nevertheless undergoes changes now and then isn't contradictory at all.
Your practical questions are all good ones, I think. I also agree that we could generate best-practices hypotheses, at least if we had a method to find optimal solutions to arbitrary problems. Unfortunately, we do not, and in fact genetic algorithms are often used in lieu of a better method; they are known to have flaws, i.e. to not be best, in a variety of situations. So while this is relevant in theory, in practice I think the method of finding a best practice deserves further attention. And you may suggest that these hypotheses are not compatible with Darwinian theories...but then we have to ask, which one better corresponds to what we observe?
(Also, why is the word "topology" in there at all? Should it, strictly speaking, be "metric space"? Or is this phrase using the non-mathematical definition of "topology" to suggest nonlinearity?)
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 30. April 2003 07:57
RBH,
Using the dynamic best practices topology there is a clear distinction between evolution as a permanent and universal process and evolution as a dynamic or evolving process. Using the dynamic best practices topology the two types of evolutionary processes are incompatible. I don’t know what type of topology or logical structure modern evolutionary biologists or geneticists use to reconcile the two forms of evolutionary process but I would be interested in seeing it defined.
What is being proposed here is a topology, a logical structure, an abstract mathematical universe with certain complex structures and principles. The topology defined, it is suggested, provides a logical mathematical structure which can be used for formulating useful predictive scientific models and hypotheses addressing the operation of life forms and systems with information processing behavior similar to that of life forms.
The topology, it is suggested, can be used not only to analyze evolutionary processes, but also complex human social behavior such as ‘the behavior of scientific analysis’.
Rex,
It could be argued that biological information processing ‘should’ follow dynamic best practices because that is the standard or approach most likely to result in survival. But this brings to the question ‘why survive?’. In my opinion, it is more appropriate to define ‘purpose’ as a useful feature of the logical/mathematical structure or topology. The use of the concept of task, function or purpose makes it possible to formulate predictive scientific hypotheses of dynamic, progressive processes. This is similar to the use of the concept of permanent and universal causation in formulating theories in physics.
The criteria for determining the acceptability of mathematical structures or topologies are 1)logical soundness/logical consistency and 2)usefulness. There is no requirement or need to show that ‘evolution should follow best practices’.
I would agree that there would be no need for the dynamic best practices topology if the ‘evolution of evolutionary processes’ has produced only minor enhancements. My contention is that the evolution of evolutionary process has, and continues to, produce some very substantial ‘improvements’ in evolutionary processes. Humans, IMO, represent a major and dramatic change in evolutionary processes( as did the development of multi-cellular organisms and the development of nervous systems).
There is a practical method of finding best practices solutions, it is called successive approximation. You start with some process or algorithm A1 and compare it to some alternative A2. If A2 is better than A1 then it becomes the base for comparison to A3. One of the key feature in making successive approximation work, is the existence of an objective criteria for determining if A1 performs a task better than A2. There are all sorts of known techniques for speeding up the search process.
Genetic algorithms represent a potentially interesting subject for discussing the dynamic best practices topology. Of particular interest is the question of "What exactly is the topology, logical structure, or logical/mathematical universe associated with GA models?".
The subject of this thread is the ‘dynamic best practices topology’. Biological systems are very complex. I am suggesting developing valid mathematical scientific models and hypotheses of these complex systems requires a complex logical structures of topologies. IMO, existing life sciences are limited by the lack of appropriate topologies. If, I suggest, 1)you define an appropriate topology, and 2)you develop a set of ‘best or better practices’ standards, practices and procedures for evaluating scientific theories which significantly reduce the bias in current academic science procedures then 3)you will produce and validate scientific models and hypotheses significantly different than the models and hypotheses currently accepted by the scientific community. This is rationale for introducing this topic. The subject being discussed is the proposed topology.
IP: Logged
|
|
Alix Nenuphar
Member
Member # 686
|
posted 30. April 2003 13:53
Warren:
My apologies if my questions were not clear. Allow me to try again. (And I was, indeed, referring to free will - though there is somewhat more to it than that.)
Your 'intelligent agents' are modeled using algorithms that follow 'Best Practises'. But actual intelligent agents do not always make optimal choices. Part of the reason for the failure of intelligent agents to make these choices lies in the restricted data available to them - i.e. the input data set is always incomplete, and is incomplete as a function of 'position' in your manifold. Precisely how do you intend to model this data loss?
Second, you indicated that various intelligent agents, given identical inputs, output choices, and the same algorithm would make different selections, even though the algorithm would choose only one. How will this be modeled, if your intelligent agents - presumably of a class or category - are all represented by that algorithm? In addition, I would suggest any less than optimal choice is not 'noise,' but rather the domain of behaviour that you wish to model.
I am unclear about your response to RBH: what precisely distinguishes the 'dynamics' of your model from the 'static' Darwinian model? Is it a self-correcting or self-modifying function that is built into the Best Practises algorithm?
Finally, you appear to be making the assumption that your Best Practise algorithms will always be able to find a single 'best-choice' output; this would appear (especially in the light of the incomplete information problem I mentioned above) to be implausible: would it not be better to expand your model to include multiple decision scenarios (perhaps by a paralysation function of the intelligent agent, or a random selection of choices)?
In addition, if you could provide more detail about the shape of the Best Practise algorithms, it would be easier to assist you in refining them.
[ 30. April 2003, 13:54: Message edited by: Alix Nenuphar ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 01. May 2003 06:24
Alix,
The topology being defined treats biological entities as intelligent agents, but does not suggest that intelligent agents always make optimal decisions. The topology suggests/assumes that under ideal conditions, or in the absence of noise and inefficiency, and given a fixed set of potential inputs, and a fixed set of potential output or responses, the agent will use the ‘best practices’ input-output algorithm which is defined as the algorithm most likely to result in achieving some goal such as survival.
Even using a best practices algorithm, organisms can make non-optimal decisions if the ‘optimal response’ is not included in the available output or if the input needed to make the optimal decision is not included in the input set. Furthermore, an intelligent agent has no knowledge of the future. The ‘most likely to achieve goal’ information, the information that results in the selection of the best practices algorithm, must therefor be based on past experience. If future conditions are different than past experience, then the ‘best practices algorithm’ may produce a non-optimal decision.
Recognizing that intelligent agents can and do make non-optimal decisions, I am curious at the source of your claim that non-optimal decisions are due to incomplete information. What is the basis of this claim?
If you exclude high level human decision making, most biological information processing seems to be highly efficient and very close to the best practices standard. Individual cells, it appears, make very large number of decisions [decisions of the form activate gene X or don’t activate gene X] which impact survival. The error rate in such decision making appears to be very low.
I am having a difficult time identifying an example of incomplete information being responsible for a departure from best practices information processing. I would be interested in what you would consider an example of this phenomena.
The difference between static and dynamic processes is a key feature of the proposed topology. The ‘laws of nature’ or forces described by predictive theories in physics are characterized as permanent and universal. The law of nature is assumed to be the same at all time and all locations. Traditional formulations of Darwin suggest or suggested that ‘evolutionary change processes’ are, or can be modeled and defined as permanent and universal processes. An alternative view is that evolution is a dynamic(and progressive) process which takes different forms at different times. The forces of evolutionary change, from this perspective are temporary and local rather than permanent and universal. The ‘computing capacity’ of computers would be an example of a dynamic and progressive process or force. Computing capacity can obviously be different for different machines and even for the same machine at different times. Evolutionary processes, at least using the dynamic best practices topology, can not be both ‘permanent and universal’ and ‘dynamic and progressive’.
The proposed topology assumes/defines that for a given time and location, and for a given set of possible input-output algorithms Fxt, and for a given task, function, goal, or purpose G, there is a member f of Fxt which is most likely to result in G. The ability of f to select the best option, does not depend on the form used to express f. I am not suggest some special type of processing algorithm which always picks the best option.
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 01. May 2003 20:55
Warren, I've suddenly become confused. Is dynamic best practices supposed to be something we apply to our practice of doing science, in order to do science better; or something we apply to evolution, using the results of d.b.p. to predict what we would find in the real world and explain how things have happened?
[ 01. May 2003, 20:55: Message edited by: Rex Kerr ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 02. May 2003 07:20
Rex,
The dynamic best practices topology is an abstract mathematical space containing ‘topological features’ which are useful in modeling and analyzing living systems. Specifically, the abstract space or universe contains entities which 1)perform information processing and 2)interact with and/or extract information from an external environment.
The information processing or biological information processing performed by the entities in this space is defined as or assumed to be dynamic or programmable. If 1)the programming or logic governing information processing in an entity is dynamic or changeable, and if 2)we don’t have direct knowledge of what program will be used at some future point in time, then 3)we can not predict how the entity will process information, how the organism will behave or what responses it will generate.
The dynamic best practices principle is an assumption or definition or convention which makes it possible to formulate and express predictive scientific hypotheses for dynamic or changeable biological information processing. If you know 1)the set of possible processing algorithms, 2)the task, function, goal or purpose of the information processing, and you assume 3)the dynamic best practices principle, then you can predict future information processing.
It will be noted that in order to formulate a predictive scientific hypotheses in traditional science, you must assume or define that causal relationships are permanent and universal. The dynamic best practices principle serves exactly the same role in formulating scientific theories as the permanent and universal or deterministic principles/assumption/definition. The dynamic best practices topology suggests that we can produce useful, reliable, testable, predictive scientific hypotheses explaining the behavior of biological systems, if we assume/define that life forms and biological information processing is purposeful or teleological.
The acceptability of the dynamic best practices principle in scientific analysis depends on two criteria- 1)is it logically sound, and 2)does it produce useful results. One of the goals of the discussion here is to evaluate the logical soundness of the proposed principle. The question of ‘usefulness’ involves issues which cannot be addressed in this forum.
Back to your question. Both evolutionary change processes and scientific analysis are manifestations of biological information processing. The dynamic best practices topology and the dynamic best practices principles can be used in the scientific analysis of both.
Using the proposed topology/principle/assumption, evolution is viewed as a dynamic process. A ‘predictive hypothesis’ of evolutionary processes would involve 1)identifying the goal of evolution as survival, and 2)descriptions or models of the change processes available to support this goal. To test such an hypothesis, we could generate and test predictions that an organism or species would or would not survive particular sets of environmental changes.
The dynamic best practices approach can also be used to construct, compare and test hypotheses regarding the performance or behavior of scientific analysis. A predictive hypothesis for scientific analysis could take the form- 1)one of the goals or tasks of scientific analysis is the formulation and validation of predictive theories, and 2)X is the best practices set of procedures for accomplishing this goal. As I described earlier, two competing theories could be offered in this format, one proposing current academic science standards as best practices and a competing hypothesis proposing some competing set of procedures as best practices. By comparing the hypothesis formulated, validated and rejected by the two sets of scientific procedures, it would be possible to determine which of the two competing sets of procedures were closer to best practices.
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 02. May 2003 10:34
One quick note: warren wrote quote:
Back to your question. Both evolutionary change processes and scientific analysis are manifestations of biological information processing. The dynamic best practices topology and the dynamic best practices principles can be used in the scientific analysis of both. (Emphasis added)
There seems to be an incipient infinite regress here. The argument must assume what it is to show in order to be used. That is, the stated goal is to examine competing modes of scientific analysis using one of those modes. That seems to presuppose the validity of the mode being used to examine itself in competition with other modes. Are there criteria that justify using "dynamic best practices principles" in such a self-analysis that are independent of those principles? On what criteria are the hypotheses mentioned in the last sentence to be compared?
RBH
[ 02. May 2003, 10:37: Message edited by: RBH ]
IP: Logged
|
|
|
Printer-friendly view of this topic