Theories, Models and Applications What is a good explanation of a complexly organized system? John Collier, Philosophy, UKZN Http:

Theories, Models and Applications

What is a good explanation of a complexly organized system?

John Collier, Philosophy, UKZNHttp://www.ukzn.ac.za/undphil/collier

Theories, Models and Paradigms: Theories

• A theory is an formal structure, together with paradigmatic applications (exemplars) and a set of preferred analogies for extensions to new applications.

• A simple example is the theory of harmonic oscillators, which has a force law defined within mechanics, and an exemplar in the pendulum. The preferred analogies are the application of sinusoidal motion governed by specific forces laws, e.g., Hooke’s law, radiation laws, etc.

• Theories are typically nested. For example the theory of harmonic oscillators is nested in a theory of mechanics, which is nested in a kinematic theory. The theory of electromagnetic radiation is a special case of the theory of harmonic oscillators. The top theory in the nesting is typically very general, if not universal.

Theories, Models and Paradigms:Models

• A model is a specific application of a theory, known as an interpretation of the theory. Each exemplar is a model showing how the theory is applied. Preferred analogies determine the scope of intended applications of the theory.

• A theory is the defined by the set of intended applications (intended models).

• For any given system it is an open question whether or not it is within the intended models of a theory. This is relatively trivial if the system in question is derived as an application of the theory.

• However, no real system can be derived, as its hypothetical nature is always subject to refutation. The a real system is a T-system is a theoretical hypothesis. The claim is that the system is a model of the theory. To be empirical, this claim must be testable.

• So it is trivial that all electromagnetic radiation systems are harmonic systems, and are mechanical systems, but it is not trivial that a particular real system is an electromagnetic system. This must be shown.

Theories, Models and Paradigms:Paradigms

• Paradigms are broad approaches using common exemplars and preferred analogies. Typically they are “top” theories.

• A typical paradigm is a Newtonian system. Newton’s hypotheses were that various particular systems are Newtonian systems, but he also boldly conjectured that all physical systems are Newtonian systems. This was strongly confirmed, but any real physical system that is not a Newtonian system (say and Aristotelian system, or an Special Relativity system) would falsify this claim. Some systems are measurably found to be SR systems, so Newton’s conjecture has been falsified.

• There is however a core of Newtonian theory characterized f=ma that has never been falsified. It could be falsified if we were to find, say, an Aristotelian physical system.

• Newton’s second law (f=ma) serves as a paradigm for all physics, guiding all studies of dynamics (the study of forces and flows).

• However there are more specific paradigms within this paradigm, such as the mechanical paradigm (all systems can be fully analyzed in terms of contributing mechanical processes by finite methods). This is best characterized by Laplace’s demon as an ideal limit. Complexly organized systems prove the mechanical hypothesis to be false.

Explanation, Confirmation and Testing: Explanation

• An explanation is to bring a real system or set of real systems under the intended interpretations of a theory. This is inevitably partially open, since the match by preferred analogies to exemplars is somewhat flexible.

• It is also somewhat open due to inaccuracies or limitations of measurements, and to interfering factors (Aristotelian accidents) that cannot be controlled for in practice.

• Measurement can be improved, and interfering factors can be controlled for, but not perfectly. For some sorts of systems, such as statistical systems or complexly organized systems, there are theoretical limits on accuracy of fit.

Explanation, Confirmation and Testing: Confirmation

• A hypothesis that a specific system S is a T-system is formed by abduction, which is the weakest form of scientific reasoning. The form is:

All the balls in this bag are black

This ball is black

Therefore, this ball is from the bag

• So we need to confirm the hypothesis. This requires that we check that S fits T closely enough to justify the hypothesis.

Explanation, Confirmation and Testing: Testing

• The hypothesis that S is a T-system can be confirmed, which is good, but it does not justify calling S a T-system in itself. We must also test the hypothesis.

• For example, von Danniken argued that much evidence confirmed the hypothesis that ancient astronauts came to Earth. Like abduction, confirmation is too easy. Many good scientists for get this.

• We must test our hypotheses in order to make them more certain.

Explanation, Confirmation and Testing: Testing 2

• A good test requires that if T is true, then S, and that if T is not true, it is highly likely that not S. The conditions can be modified for statistical predictions.

• A good explanation is one that has been tested and passed. A potentially good explanation is one that can be tested. All other explanations are bad explanations.

• Inference to the best explanation applies only to good explanations. There is no point in finding the best explanation among a set of bad explanations. It is still a bad explanation.

Testing NeoDarwinist Hypotheses

• NeoDarwinism is the hypothesis that all biological systems are neoDarwinian systems (produced by selection through gradual processes on the canonical form, though exceptions are know to the graduality and selection conditions, but these are theoretically acceptable within limits of the mechanisms supposed – e.g., adaptive radiation following major structural change bottlenecks, genetic drift).

• NeoDarwinist hypotheses, to be good, must be such that given neoDarwinism the hypothesis should be true, and if not neoDarwinism, the hypothesis should be highly likely to be false.

• Darwin’s finches are a classic case. Darwin made a hypothesis that the finches’ phylogeny is a neoDawinian system, and showed the evidence confirms this. More recent studies have strengthened the confirmation, but have also tested the hypothesis with genetic information, making it highly improbable that the distribution is not due to genetic change. Further testing could be done with knowledge of the specific genes involved and their relations to different phenotypes. Note that the hypothesis is falsifiable, and some finches on some islands of the Galapagos may not fit.

Testing NeoDarwinist Hypotheses: When neoDarwinists go wrong

• As mentioned, an abduction to a confirmed hypothesis does not necessarily produce a good hypothesis. We must test the hypothesis, or at least ensure it is testable.

• Optimality theorists typically make the mistake of not requiring tests. A typical argument goes: Trait t would exist if it were selected for optimization of some value important to survival (e.g., copulation time in fleas). There is typically no attempt to show either that optimization or selection occurred, since this is presupposed in the model. But the model is the very thing in question, so this argument is circular.

• Typically, though, the hypothesis can be tested, most because the mechanisms at work are of known kind, and the explanations are mechanistic, and hence mathematically closed by finite methods.

Emergence Explanations 1

• Emergence and self-organization are closely linked, and I will assume that spontaneous self-organization is the source of emergent structures and processes.

• Organization can appear from previous organization, or through emergence. In some cases the first case results from the following of instructions, but in others it results from natural organizing processes that remove nonorganized aspects of the system. Traditional views of development and heredity fit the instruction model. The formation of oil globules and fatty membranes fits the latter case, which we call self-reorganization. Self reorganization requires the dissipation of information to leave the organized part. This is not emergence, as no novelty is formed.

Emergence Explanations 2• Emergence explains new organization as a production of self-organizing

processes (e.g., Kauffman, though he does not carefully distinguish between self reorganization and self-organization).

• Traditional properties of emergent forms (Weber talk) should appear in systems explained via emergence (especially novelty, nonreducibility and nonpredictability). Likely appearance of those properties confirms emergence explanations. They must still be tested.

• Robert Rosen (Life Itself) gave conditions for emergence (nonmechanism) in terms of a) properties of models and b) properties of networks, and claimed these are equivalent. The network property implying emergence is closure to efficient causation. However this is not a readily observable condition, since there may well be missed open network descriptions for a network with closed loops.

• So models using emergence explanations must be tested independently. Otherwise we can only invoke the fitness (confirmation) of the model to the real systems in question.

• However, since predictability and reducibility fail in emergence explanations, how to do this is not straight forward.

Can we test emergence explanations?

• The basic problem is that we can confirm emergent explanations, but that is not enough; we need to test the explanations.

• The answer to the testing problem, as in the case of selection theory, is that we must pay attention to the dynamics of the hypothesized model, and see whether or not the dynamics of the actual system support the model or not. The problem is that we cannot in this case test directly, since predictability and reduction fail.

• The best we can do is to test for the right sort of dynamical conditions in the system for the hypothesis to be true.

• Why dynamical conditions? Because only dynamical conditions have effects, and can be controlled. Anything that involves no dynamical conditions is not subject to manipulation, and is not accessible to us through experimental method (or observation).

• Typically we can either determine the dynamical properties of an emergent system by looking at the basic forces involved by the type of objects involved, or by taking the system apart and examining the sorts of forces involved, assuming that the forces do not mysteriously change in the assembled system. In many cases this works.

Testing emergence: example 1

Solar system harmonics1. Moon2. Jupiter’s moons3. MercuryChaotic regions in phase space are resolved through

dissipation (self-reorganization) to form harmonic orbits. Selection of orbit is nonreducible and nonpredictable. The dynamics are Lagrangian, but nonHamiltonian.

Lesson: dissipation is required for self organization; the system must be nonHamiltonian, and not a step function.


Bénard cell convection

To solve, we take the equations of motion of the convecting state and the nonconvecting state and set them equal to get the transition point. This turns out to be the point of minimal entropy production. In other cases we get a minimization of entropy production in the direction of applied forces (Kay and Schneider, Prigogine).

Lesson: Minimization of entropy production (friction) is typical of self-organization – it is intrinsically tied to system efficiency. Self-organization resists external forces and contributes to system independence and individuation.

Testing emergence: example 3Network emergence in the Sea Urchin A Genomic Regulatory Network for Development• Eric H. Davidson, Jonathan P. Rast, Paola Oliveri, Andrew Ransick, Cristina Calestani, Chiou-

Hwa Yuh, Takuya Minokawa,1 Gabriele Amore,1 Veronica Hinman, Cesar Arenas-Mena, Ochan Otim, C. Titus Brown, Carolina B. Livi, Pei Yun Lee, Roger Revilla, Alistair G. Rust, Zheng jun Pan, Maria J. Schilstra, Peter J. C. Clarke, Maria I. Arnone, Lee Rowen, R. Andrew Cameron, David R. McClay, Leroy Hood, Hamid Bolouri

1 MARCH 2002 VOL 295 SCIENCE www.sciencemag.org pp. 1669-1678• Abstract Development of the body plan is controlled by large networks of regulatory genes. A

gene regulatory network that controls the specification of endoderm and mesoderm in the sea urchin embryo is summarized here. The network was derived from large-scale perturbation analyses, in combination with computational methodologies, genomic data, cis-regulatory analysis, and molecular embryology. The network contains over 40 genes at present, and each node can be directly verified at the DNA sequence level by cis-regulatory analysis. Its architecture reveals specifc and general aspects of development, such as how given cells generate their ordained fates in the embryo and why the process moves inexorably forward in developmental time.

• “But even from the first-stage model, But even from the first-stage model,which just states the interactions that occur at each node, there emerge system properties that can only be perceived at the network level. which just states the interactions that occur at each node, there emerge system properties that can only be perceived at the network level.”

• This suggests that there are emergent properties in the network, and would be a confirming instance. Is there a test of this hypothesis?

http://www.sciencemag.org/

Testing emergence: example 3• How might these network properties be produced in development? • Assembly of parts from earlier subnetworks, assembled themselves from

smaller genetically translated parts. This seems not to be the case, given the stages mapped and their dynamical relationships.

• Self reorganization. This is possible on the evidence in the paper, with constructing parts being dissipated as development progresses.

• Spontaneous self organization. This is also possible on the evidence in the paper. As above producing processes dissipate some of the order.

• In order to distinguish between reorganization and spontaneous self organization, further studies are required. But the case is a candidate, and the test could be on the mapped processes with further analysis for whether there is the production of new information, or all information is in the original protoplasm and genetic information.

• Similar considerations apply to the evolutionary origin of the network properties, but testing will require comparison across other families (all sea urchins have the same first 16 stages).

Testing emergence: example 4A Complex Tail, Simply Told, Science Now, April 17, 2007. Bacterial Flagellae• Ochman and Liu report their findings online this week in Proceedings of the National Academy of

Sciences: Renyi Liu, and Howard Ochman Stepwise formation of the bacterial flagellar system PNAS 104: 7116-7121; published online before print as 10.1073/pnas.0700266104

• A suite of more than 50 genes builds and operates the flagellum.• Evolutionary biologist Howard Ochman and postdoc Renyi Liu of the University of Arizona,

Tucson, obtained the complete genomes of 41 flagellated bacteria species and identified 24 flagella-related genes common to all the microbes.

• In each species, the 24 genes were very similar to each other but not to any other genes in the genome.

• This finding, coupled with the observation that this complete set of genes exists in all flagella-bearing bacteria, suggests the genes arose by duplication of a single gene in the ancestor of all bacteria.

• Slight changes in the genes then generated new functions. Each gene is responsible for a different aspect, such as producing the proteins that make up the flagellar motor, filament, and other structural components.

• In addition, an evolutionary tree constructed by the researchers suggests that the order in which the genes appeared matches the sequence of steps in the assembly of the flagellum.

• It also provides a straightforward counterexample to claims from "intelligent design" proponents that the flagellum could not have evolved from a single gene, adds cell biologist Ken Miller of Brown University.

Copyright ©2007 by the National Academy of Sciences

Liu, Renyi and Ochman, Howard (2007) Proc. Natl. Acad. Sci. USA 104, 7116-7121

Fig. 1. Distribution of flagellar proteins (excluding chemotaxis proteins) among flagellated bacterial species



Fig. 2. Congruence between species tree and flagellar protein tree



Fig. 3. Network of relationships among flagellar core proteins



Fig. 4. Protein sequence similarity among the proximal rod protein FlgF, the distal rod protein FlgG, and the hook protein FlgE in E. coli


• Does the explanation require some form of self organization?

• The evidence is not sufficient for emergence. It is compatible, but not testable for emergence or self-reorganization.

• The hypothesis that it is due to gradualist selection on large changes in the genome die to duplication is possible, and is the current best explanation. Further testing is required for this hypothesis.

Some other examples of self organization in biology

• Ant behaviour (nest building and bridge building) (widely discussed)

• Differentiation of meristems in early development of plants (Banerjee and Collier)

• Sympatric Speciation models (Collier and Banerjee)

• Various cases from Stu Kauffman (his talk)• So far these hypotheses vary from speculative

through confirmed to tested. Testing along the lines suggested needs to be applied more widely.

Conclusions• Organized complexity comes from organized complexity or from self organization.• Self organization models have been used, and are in various degrees of confirmation

and testing.• Organized complexity coming from pre-existing should be traceable to pre-existing

organized complexity. This is required for testing the hypothesis of pre-existing organized complexity as the source of later organized complexity (both in development and evolution).

• Darwinian selection supposes complexity arises though the accumulation of environmental influences on greater complexity produced by organisms. It does not require the preexistence of organization, which appears by chance accumulation. However, sufficient pre-existing complexity must exist in the environment to account for the accumulation of functional complexity.

• Reorganization requires pre-existing complexity of the same order, and the organization of the complexity implicit must be as high as the resulting organized complexity – only “noise” is removed by dissipation.

• Self-organization does not require either pre-existing complexity or organization, but does require self-organization conditions.

• All of the required conditions can be determined by observations, so potentially it is possible to test among good version of the relevant hypotheses.

Documents

Theories, Models and Applications What is a good explanation of a complexly organized system? John Collier, Philosophy, UKZN Http: