Brian Miller: Thermodynamics and the Origin of Life

Office Hours
2

I don’t have the time to read all the articles now, so I will comment only on the first article, “Thermodynamics of the Origin of Life”.

My understanding from talking to my astrobiologist colleagues is that the statistical unlikeliness of life forming is indeed a puzzle. I am not an astrobiologist myself, so my views are somewhat simplistic, but the main problem as I see it is the following:

In a thermodynamics process that could create either a living or a nonliving thing, the ratio of probabilities are given by the ratio of Boltzmann factors:

P(living)/P(nonliving) = (g_living/g_nonliving)*Exp[-(E_living - E_nonliving)/kT)]

Where E_living is the energy of a living thing, E_nonliving is the energy of a nonliving thing, g_living the number of states that can be considered alive, and g_nonliving the number of states that is considered nonliving, k is the Boltzmann constant, and T is the temperature of the system in which these processes occur (i.e. temperature of the primordial muck).

Putting it back to the language he used in the article:

  1. The Exp[-(E_living - E_nonliving)/kT)] term is the one that suppresses the possibility of living things due to living things having a larger energy than nonliving things (E_living>E_nonliving). I don’t know if that is true, but it seems sensible.
  2. The (g_living/g_nonliving) term is where entropy enters the equation. Supposedly there are more states that we would call living than nonliving, so g_living/g_nonliving<1. This illustrates that higher entropy states (higher g) is more likely to occur than lower entropy states.

Here are my issues with the statement that “Science Show Intelligence Was Required” for life:

  1. First, a lot of these terms are unknown. There are many states that could be considered living, and many states that could be considered nonliving, but in the end the ratio g_living/g_nonliving is unknown. I don’t think that the space of allowable states is even known, which makes it hard to properly evaluate g_living and g_nonliving. Further, these different states have different E_livings and E_nonlivings. The point is that no rigorous calculation has been made to compute P(living)/P(nonliving). I understand that P(living)/P(nonliving)<1, but how small is debatable.
  2. Next, to get the ratio of the number of living things/number of nonliving things, P(living)/P(nonliving) needs to be multiplied by the number of interactions that could produce living things/nonliving things. I am not sure how many interactions can happen in a span of time - this is a question for the biologists, but the number could be very large. Even if P(living)/P(nonliving) is small, if the number of interactions is large, it is still possible to produce living things.
  3. It is not true that adding energy to the system does not help. If the energy (say sunlight shining into a primordial pond filled with muck) heats the system, increasing the temperature of the system, then P(living)/P(nonliving) is going to be driven closer to its maximum value of g_living/g_nonliving. To see this, set T->infinity in the equation above.
  4. There is the whole debate about non-equilibrium thermodynamics in the article, which I will skip because I know next to no non-equilibrium thermodynamics.
  5. Moving to more philosophical objections: this is a God-of-the-gap argument. Just because right now it is a puzzle that life is statistically unlikely, does not mean that in the future we won’t find a mechanism that explains this neatly.
  6. I don’t think something that pushes P(living) up has to necessarily be equated to an “Intelligence”. Might be some non-random, but thoughtless thing.
4 Likes