On Uncommon Descent the erroneous idea that the second law of thermodynamics contradicts evolution is still being peddled (See here and here). The reason why the second law is consistent with evolution is not difficult to understand. Trouble is, the passionate anti-evolutionists have a strong contingent of Young Earthers in their midst. The latter, going right back to the days of Morris and Whitcomb’s book “The Genesis Flood”, have committed themselves to this fallacy and now they cannot budge from it without becoming a laughing stock. I have tackled this subject several times on this blog, but there is no harm dealing with the subject again from a slightly different angle.

In physics the entropy of a physical system given its stipulated macroscopic conditions is an increasing function of the number of microscopic states consistent with those conditions - the latter number is referred to as the statistical weight of macrostate. For example, if we stipulate that the system is a gas with stated pressure, temperature and density then the entropy will be determined by the number of ways such a system can be realized. If we stipulate that the system is crystalline then it is fairly obvious that the entropy of such a system is relatively low as there are, relatively speaking, not many microstates that realize such a configuration. More abstract macroscopic conditions can be conceived such as requiring that the system is proactively self perpetuating as in the case of living structures. Interestingly, it is clear that living structures are far more “disordered” than crystals in as much as the number of ways it is possible to exist as a proactively self perpetuating configuration are myriad compared to the number of ways a crystal structure can exist. Entropy as a quantity does not pick up on the fact that between the extreme ends of its rather undiscerning spectrum, there are remarkable structures that from an entropy point of view are not differentiated as particularly special. Another limitation of entropy as a quantity is that it only measures the statistical weight of macrostates given the constraints of the physical regime and these meta-constraints remain untouched by thermodynamic decay. Entropy is therefore not a measure of absolute disorder: It is possible to conceive of physical rĂ©gimes that are so restricted that only highly ordered configurations are consistent with those regimes; in such a context entropy only provides a measure of those macrostates with the greatest statistical weighting. Thus increasing entropy does not necessarily imply a decay to absolute disorder.

That the second law of thermodynamics is not inconsistent with evolution becomes clear even if we start by assuming that the cosmos is the product of an intelligent designer. In fact, what now follows is an intelligent design argument against the YEC abuse of the second law of thermodynamics. If our intelligent designer is endowed with the level of omnipotence and omniscient intelligence usually associated with the Judeo-Christian deity, then given a construction set of parts such an entity would be able to conceptually assemble the entire space of configurational possibilities open to that set of parts. This configuration space forms a kind of manifold of points and a Judeo-Christian deity would be able to think about this manifold like a human being thinks about 3D space. As it stands, however, this huge platonic object is pretty dead and static. So the next step is to introduce some kind of physics in order to give it dynamism. To this end our Judeo-Christian deity could proceed to “wire” up this configuration space into a network of connections. Most naturally the metric of this network would recognize the fact that the manifold of configurations naturally forms a network of relations: These natural relations exist by virtue of the spatial relations resulting of the fact that some configurations are only separated by a small distance in terms of the incremental adjustments needed to turn one into the other.

So, now we have a manifold of points connected into a network by some kind of connection metric. But we still have a pretty static object. The next thing is to give this network a dynamic by assigning transition rates to the connections: This means that if the system is known to be in one configuration we can then work out the probability of it making a transition to one of its “nearby” configurations. The manifold now has a dynamic; that is, it has some physics: Given these transition rates the system will now move from one configuration to the next.

Let us represent this dynamic by the function, T(L), that maps the links represented by L to corresponding transition rates T. In its most straightforward form T will consist simply of a list that maps the links between the nodes of the manifold to their respective transition rates. Given the powers of our assumed deity, then it is clear that such a being has available to it an enormous number of choices on how to wire up this network and how to assign transition rates: In particular it would be quite within the powers of this deity to so wire up this network that it would move toward configurations that contain living structures. To achieve this the transition probabilities need not be directionally biased: The function T may form narrow channels of flow where no direction is favoured, but because these channels form such narrow bottle necks the configurations containing life would have relatively high statistical weight thus considerably enhancing the probability that such configurations arise as the system moves through configuration space. If within the specified channels of development the transition rates are isotropic then this implies that the motion within these channels is one of unbiased random walk. From random walk immediately follows the second law of thermodynamics; namely, that the system would tend to move toward those macrostates with the greatest statistical weight - which is all the second law tells us. Since T(L) puts a tight restriction on what is possible, the macrostates with the greatest statistical weight (that is, with the greatest entropy) are not necessarily disordered in absolute terms. Thus configurations containing living structures can evolve and yet the second law not be violated. The second law works within the constraint supplied by the function T(L) whose form is not subject to thermodynamic decay and is selected by divine intelligence to considerably enhance the probability of life arising. Here then is the rub for those who naively think the second law to contradict evolution: The above system would simply migrate towards its most probable macrostate, that is the macrostate with the greatest “disorder”, and yet if T(L) is carefully chosen by our super–intelligent deity these so called "disordered" states may contain what in absolute terms are the highly complex ordered configurations of living structures.

It all comes down to how the function T(L) maps transition rates to the links between configurations. There is actually nothing really profound here: Given the freedom in choosing any arbitrary T(L) Divine Intelligence is quite capable of contriving a network of transition rates in such way as to favour the evolution of life. The function T(L) effectively defines the physics of the system; that is, it tells us the probability of a system moving for one state to the next. However, it is a funny sort of physics as it simply takes on the form of a list of connections and associated transition rates. This list of connections will contain a high level of information on two counts:

a) It will be a very long list and thus in terms of its linear size it will contain a lot of information.

b) It will be simply one list of many, many possible lists and it will therefore be an extremely rare selection. If, assuming equal a-priori probabilities, we equate this rarity to a selection probability then the implied improbability will entail a very high level of information as defined by the expression for information, –log P. (Which is the expression ID theorist William Dembski uses)

But the profound and difficult questions are these: Is it possible to compress and encode the information in this list into a set of elegant laws? In fact, is our system of physical laws one such compression? I’m not sure I know the answer to these questions, but they are questions I have never seen properly framed let alone addressed on Uncommon Descent or in any of the papers I have read written by anti-evolutionists. Instead the anti-evolutionist stance has a tendency to encourage a spurious dichotomy between “naturalism” and intelligent design. Naturalism is the view that somehow elemental nature can go it alone, a view which is at the heart of atheism. However, no doubt unintentionally, the views expressed by the anti-evolutionists appear to promote the concept of naturalism: The anti-evolutionists who follow William Dembski loudly proclaim the virtues of their design detection science oblivious to fact that it is easy to construe this as suggesting that some things in nature are “designed” and therefore “artificial” and some things need no design and therefore are “natural” , uncreated by intelligence.* I’m sure that in their heart of hearts Dembski and his followers don’t intend this insinuation, but it is all too easy to read the anti-evolutionist thesis as setting up a dualist category of nature versus God. They have created a PR problem for themselves and this is indicated by the fact that Dembski feels the need to address it here.

The fallacious use of the second law by the anti-evolutionist lobby only serves to reinforce this false dichotomy between nature’s creative power and Divine creative power: Thus the anti-evolutionists who have an deep instinctual fear of evolution feel the need to have ready a killer proof of the superiority of Divine creative power over the much feared apparent creative power nature. But the truth is that the second law is no killer argument against evolution and in any case the apparent power of nature to create via evolution must ultimately trace back to the Divine ability to specify the function T(L).

Footnote

* Some problems are harder than others: The presence of a solution to a difficult problem may give an indication of the level of intelligence that solved it.

In physics the entropy of a physical system given its stipulated macroscopic conditions is an increasing function of the number of microscopic states consistent with those conditions - the latter number is referred to as the statistical weight of macrostate. For example, if we stipulate that the system is a gas with stated pressure, temperature and density then the entropy will be determined by the number of ways such a system can be realized. If we stipulate that the system is crystalline then it is fairly obvious that the entropy of such a system is relatively low as there are, relatively speaking, not many microstates that realize such a configuration. More abstract macroscopic conditions can be conceived such as requiring that the system is proactively self perpetuating as in the case of living structures. Interestingly, it is clear that living structures are far more “disordered” than crystals in as much as the number of ways it is possible to exist as a proactively self perpetuating configuration are myriad compared to the number of ways a crystal structure can exist. Entropy as a quantity does not pick up on the fact that between the extreme ends of its rather undiscerning spectrum, there are remarkable structures that from an entropy point of view are not differentiated as particularly special. Another limitation of entropy as a quantity is that it only measures the statistical weight of macrostates given the constraints of the physical regime and these meta-constraints remain untouched by thermodynamic decay. Entropy is therefore not a measure of absolute disorder: It is possible to conceive of physical rĂ©gimes that are so restricted that only highly ordered configurations are consistent with those regimes; in such a context entropy only provides a measure of those macrostates with the greatest statistical weighting. Thus increasing entropy does not necessarily imply a decay to absolute disorder.

That the second law of thermodynamics is not inconsistent with evolution becomes clear even if we start by assuming that the cosmos is the product of an intelligent designer. In fact, what now follows is an intelligent design argument against the YEC abuse of the second law of thermodynamics. If our intelligent designer is endowed with the level of omnipotence and omniscient intelligence usually associated with the Judeo-Christian deity, then given a construction set of parts such an entity would be able to conceptually assemble the entire space of configurational possibilities open to that set of parts. This configuration space forms a kind of manifold of points and a Judeo-Christian deity would be able to think about this manifold like a human being thinks about 3D space. As it stands, however, this huge platonic object is pretty dead and static. So the next step is to introduce some kind of physics in order to give it dynamism. To this end our Judeo-Christian deity could proceed to “wire” up this configuration space into a network of connections. Most naturally the metric of this network would recognize the fact that the manifold of configurations naturally forms a network of relations: These natural relations exist by virtue of the spatial relations resulting of the fact that some configurations are only separated by a small distance in terms of the incremental adjustments needed to turn one into the other.

So, now we have a manifold of points connected into a network by some kind of connection metric. But we still have a pretty static object. The next thing is to give this network a dynamic by assigning transition rates to the connections: This means that if the system is known to be in one configuration we can then work out the probability of it making a transition to one of its “nearby” configurations. The manifold now has a dynamic; that is, it has some physics: Given these transition rates the system will now move from one configuration to the next.

Let us represent this dynamic by the function, T(L), that maps the links represented by L to corresponding transition rates T. In its most straightforward form T will consist simply of a list that maps the links between the nodes of the manifold to their respective transition rates. Given the powers of our assumed deity, then it is clear that such a being has available to it an enormous number of choices on how to wire up this network and how to assign transition rates: In particular it would be quite within the powers of this deity to so wire up this network that it would move toward configurations that contain living structures. To achieve this the transition probabilities need not be directionally biased: The function T may form narrow channels of flow where no direction is favoured, but because these channels form such narrow bottle necks the configurations containing life would have relatively high statistical weight thus considerably enhancing the probability that such configurations arise as the system moves through configuration space. If within the specified channels of development the transition rates are isotropic then this implies that the motion within these channels is one of unbiased random walk. From random walk immediately follows the second law of thermodynamics; namely, that the system would tend to move toward those macrostates with the greatest statistical weight - which is all the second law tells us. Since T(L) puts a tight restriction on what is possible, the macrostates with the greatest statistical weight (that is, with the greatest entropy) are not necessarily disordered in absolute terms. Thus configurations containing living structures can evolve and yet the second law not be violated. The second law works within the constraint supplied by the function T(L) whose form is not subject to thermodynamic decay and is selected by divine intelligence to considerably enhance the probability of life arising. Here then is the rub for those who naively think the second law to contradict evolution: The above system would simply migrate towards its most probable macrostate, that is the macrostate with the greatest “disorder”, and yet if T(L) is carefully chosen by our super–intelligent deity these so called "disordered" states may contain what in absolute terms are the highly complex ordered configurations of living structures.

It all comes down to how the function T(L) maps transition rates to the links between configurations. There is actually nothing really profound here: Given the freedom in choosing any arbitrary T(L) Divine Intelligence is quite capable of contriving a network of transition rates in such way as to favour the evolution of life. The function T(L) effectively defines the physics of the system; that is, it tells us the probability of a system moving for one state to the next. However, it is a funny sort of physics as it simply takes on the form of a list of connections and associated transition rates. This list of connections will contain a high level of information on two counts:

a) It will be a very long list and thus in terms of its linear size it will contain a lot of information.

b) It will be simply one list of many, many possible lists and it will therefore be an extremely rare selection. If, assuming equal a-priori probabilities, we equate this rarity to a selection probability then the implied improbability will entail a very high level of information as defined by the expression for information, –log P. (Which is the expression ID theorist William Dembski uses)

But the profound and difficult questions are these: Is it possible to compress and encode the information in this list into a set of elegant laws? In fact, is our system of physical laws one such compression? I’m not sure I know the answer to these questions, but they are questions I have never seen properly framed let alone addressed on Uncommon Descent or in any of the papers I have read written by anti-evolutionists. Instead the anti-evolutionist stance has a tendency to encourage a spurious dichotomy between “naturalism” and intelligent design. Naturalism is the view that somehow elemental nature can go it alone, a view which is at the heart of atheism. However, no doubt unintentionally, the views expressed by the anti-evolutionists appear to promote the concept of naturalism: The anti-evolutionists who follow William Dembski loudly proclaim the virtues of their design detection science oblivious to fact that it is easy to construe this as suggesting that some things in nature are “designed” and therefore “artificial” and some things need no design and therefore are “natural” , uncreated by intelligence.* I’m sure that in their heart of hearts Dembski and his followers don’t intend this insinuation, but it is all too easy to read the anti-evolutionist thesis as setting up a dualist category of nature versus God. They have created a PR problem for themselves and this is indicated by the fact that Dembski feels the need to address it here.

The fallacious use of the second law by the anti-evolutionist lobby only serves to reinforce this false dichotomy between nature’s creative power and Divine creative power: Thus the anti-evolutionists who have an deep instinctual fear of evolution feel the need to have ready a killer proof of the superiority of Divine creative power over the much feared apparent creative power nature. But the truth is that the second law is no killer argument against evolution and in any case the apparent power of nature to create via evolution must ultimately trace back to the Divine ability to specify the function T(L).

Footnote

* Some problems are harder than others: The presence of a solution to a difficult problem may give an indication of the level of intelligence that solved it.