I agree the second law gets abused more often than it should…but it also gets dismissed more often than it should. It is possible that this will be review for most, but the current thread makes me think not for all.
The basis of the second law is the concept of multiplicity. Let’s think about a warm cup of coffee cup sitting on the counter away from any heat source. Since the air in the room is cooler (meaning less energetic on average) than the coffee, each random interaction has a much better chance of pairing a less energetic air molecule with a more energetic coffee molecule that it does of pairing a more energetic air molecule with a less energetic coffee molecule…so the coffee cools.
In most cases, there is enough energy in the room to heat the coffee, so it is not impossible that the coffee could warm randomly…there is no physical law that prevents it. However, if we had a way to examine every possible interaction in the system we would find there are MANY more interactions that would cool the coffee than there are interactions that would heat it. The law of large numbers acting on the imbalance in the probabilities almost guarantees that if the interactions are left to chance the coffee will cool.
In the case if hair and fingernails, the physiology of a living human body actually reverses the inherent probabilities between growth and decay. If the body is working correctly, the chances of hair and fingernail growth is a near certainty, but that does not mean that the second law is suspended…it just means there is work being done to overcome the normal probability imbalance that greatly favors disorder. Photosynthesis is similar…cell has information and hardware that is capable of doing the work necessary to keep the probabilities balanced in the plant’s favor.
Eventually the systems ability to do that work will cease (due do a lack of energy input or some decay in the system itself) and the balance will return to favor decay. We can be assured it will happen just because there are always more ways for random interactions to introduce disorder into a a system than there for random interactions to introduce more order.
The most common dismissal of the second law is the compensation argument…that the loss of order in the sun more than offsets the gains seen throughout evolutionary history…is too often uncritically accepted. The disorder increasing in the sun is essential as an energy source…but the raw energy and heat it provides to the system is not sufficient in itself to balance the probabilities involved in evolutionary scenarios. For order to win, something has to do the real work of bringing the probabilities involved more into balance.
Since there is positive evidence in many scientific disciplines that evolution has occurred, we have scientific reasons to think that the whole of the evolutionary process is capable of balancing the probabilities. At the same time I think it is fair to say that people who believe that evolution has occurred usually are convinced more by that positive evidence than they are by a thorough analysis of the multiplicity involved in how the various probabilities balance.
Personally, I don’t believe selection alone working on truly random (meaning unguided in some way…natural or supernatural) is enough to overcome the probability imbalance that most everyone admits exists…even if they argue endlessly about the degree of the imbalance. I come from a background in physics and information systems design, so I understand the complexity involved in building information processing systems (or any physical systems for that matter) that function well…and I have very little fear that an evolutionary software design system will someday take my job.
Random changes can and will move the system toward order some percentage of the time, but if any significant imbalance exists between the probability of advance and the probability of decline, the multiplicity of the system will result in a net decrease in order over time.
If evolution works…and like I already said…there is a large amount of positive scientific evidence that it does (so be kind :-))…it is because SOMETHING is causing the individual mutation events to move toward order (in the form of fitness for the environment) at least slightly more often than they move toward disorder…or the process would be powerless to resist the law of large numbers acting on the fact that each truly random unguided change has a MUCH better chance of breaking something than it does of building something.