-2

Physics tells us that universally entropy is always increasing. However, it is possible for entropy to decrease locally at the expense of a bigger increase elsewhere.

Can this happen spontaneously or does it always require intentional actions? In other words, does a local decrease of entropy require life, an organism trying to remain alive or achieve some other goals?

I have asked this question on the physics site, but I have not got any clear answer due to the fact that this question is in the grey zone between physics and philosophy.

Pertti Ruismäki
  • 1,625
  • 3
  • 11
  • Unless salty water has intent to grow salt crystals, no. See [Crystallization](https://en.wikipedia.org/wiki/Crystallization#Thermodynamic_view). – Conifold Aug 20 '21 at 10:15
  • Crystallization does not decrease entropy. – Pertti Ruismäki Aug 20 '21 at 10:18
  • Sure it does, locally. Entropy of the crystal lattice is lower than of the solution, but "thermal randomization of the surroundings compensates for the loss of entropy that results from the reordering of molecules within the system". – Conifold Aug 20 '21 at 10:23
  • Entropy is not disorder. Crystallization does not increase differences in energy density. – Pertti Ruismäki Aug 20 '21 at 10:40
  • 1
    Are you sure, Pertti? Crystallization is exothermic, and would fit your definition of "spontaneous local decrease of entropy." – ConnieMnemonic Aug 20 '21 at 10:50
  • It would have to be endothermic. After a decrease in local entropy there should be more energy available for work. – Pertti Ruismäki Aug 20 '21 at 11:01
  • 1
    This seems like a physics question rather than a philosophical one, and you can find a number of examples of non-living systems that cause localized decreases in their internal entropy by exporting entropy into their environment, see the answers to [this question on the physics stack exchange](https://philosophy.stackexchange.com/questions/4619/would-all-anti-entropic-forces-be-considered-living), including the [example of a hurricane](https://philosophy.stackexchange.com/a/4641/10780). – Hypnosifl Aug 20 '21 at 11:36
  • Physics question. And a trivial one, anyone who has seen ice forming in the winter knows entropy can naturally decrease locally. – armand Aug 20 '21 at 12:17
  • 2
    You got 3 clear answers to the question on the physics site. They just weren't what you were hoping to read. – D. Halsey Aug 20 '21 at 17:26
  • Do you think that single-celled organisms have "intentions?" – Sandejo Aug 20 '21 at 19:09
  • 2
    Please read the link or a physics textbook. With or without "disorder" crystallization decreases entropy. This is not Physics SE to explain the mechanism to you, but it looks like people there already explained several others under the [duplicate question](https://physics.stackexchange.com/q/659544/65263) you posted. – Conifold Aug 20 '21 at 21:23

1 Answers1

1

Ill-formed question. Entropy is not defined for open systems.

Physics tells us that universally entropy is always increasing.

Not precisely. a) It does not increase universally: impossible to know that; b) in closed systems, entropy increases until a maximum level, and can't increase more, it is not "always increasing".

However, it is possible for entropy to decrease locally at the expense of a bigger increase elsewhere.

Not precisely (not a bigger increment, but possibly an equivalent increment), but yes, in general terms.

Can this happen spontaneously or does it always require intentional actions?

The laws of thermodynamics (ergo, the 2nd law) are defined for closed systems, that is, for behaviors that always will occur spontaneously at the interior. Any external interference with a system cannot be described by thermodynamics. The laws of thermodynamics are empirically true, but only for closed systems. External actions (open systems) are not described by the laws of thermodynamics.

Moreover an "external action" to a system poses a whole set of conflicts to thermodynamics. Any "external action" makes the system be ontologically modified into a new one (so you would need to take the definition of the system in consideration). Here, you are falling into a common error on the interpretation of the second law: if you inject energy into a system, you are in fact modifying the entropy of the suprasystem that contains the injector, and the injected systems. You cannot calculate the entropy of isolated subsystems (try calculating the entropy change of a single molecule from a gas in a classical two-container system which division wall is opened, and you will see the problem).

Anyway, it seems you are asking yourself how does entropy increase. Remark that the laws of thermodynamics are constrained to systems that a) already exist and b) are closed. Thermodynamics do not describe systems that start to exist (having a low entropy value). Worst even, it does not describe open systems. Worst even, it is far from describing open systems that come to existence (e.g. a newborn).

RodolfoAP
  • 6,580
  • 12
  • 29
  • A decrease in entropy requires that energy is imported in the system and part of it is stored for later use as temperature differences or other forms of energy. – Pertti Ruismäki Aug 20 '21 at 11:16
  • Julian Barbour has written at some length that thermodynamics is not directly applicable to open or unconstrained systems; see his *Janus Point*(2020) or earlier https://arxiv.org/pdf/1602.08019.pdf . – sand1 Aug 20 '21 at 11:27
  • *External actions (open systems) are not described by the laws of thermodynamics* But you can model an open system interacting with some environment by drawing a larger box around the subsystem + environment, as with modeling how particular molecules will behave when in contact with a heat reservoir. Schrodinger had no problem with applying the laws of thermodynamics to living things in his book [*What is Life?*](https://en.wikipedia.org/wiki/What_Is_Life%3F) and many subsequent physicists have done the same. – Hypnosifl Aug 20 '21 at 11:33
  • @Hypnosofil: A system with subsystems is not an open system, it is still a closed system, and the entropy you are addressing is that of the whole, not that of the subsystem (try calculating what you are suggesting: as soon as you address the whole, you lose information of the parts). Otherwise, the problem of entropy for open systems would be solved. Regarding Schroedinger, his book is highly speculative. In addition, he has huge flaws in his interpretation of thermodynamics. Give it a critical read. – RodolfoAP Aug 20 '21 at 12:14
  • @PerttiRuismäki: "importing energy" into a system does not decrease its entropy. It decreases the entropy of the suprasystem (just analyze the formulae). You cannot calculate the change in entropy of a system that becomes a different thing: injecting energy into a system makes it not anymore the same system; so, you are calculating a single value of entropy for two different systems, one that existed previously, and one that existed later. That idea is quite naive. – RodolfoAP Aug 20 '21 at 12:18
  • @RodolfoAP - I didn't mean that the whole combination of subsystem + environment was an open system, but the subsystem is an open system, and since you can analyze the combined closed system using the usual methods of thermodynamics, this naturally includes the ability to predict the behavior of the subsystem. – Hypnosifl Aug 20 '21 at 15:14
  • "Moreover an "external action" to a system poses a whole set of conflicts to thermodynamics." My thermodynamics textbook was almost all about problems of changing the closed system by adding or removing heat, changing the volume, etc. So long as the changes are made slowly enough that the system is always in equilibrium, the changes fall within the scope of thermodynamics. – David Gudeman Aug 20 '21 at 19:30
  • @DavidGudeman, ..."changing the closed system by adding or removing heat, changing the volume, etc." means having two open subsystems which exchange energy (therefore, one single closed (supra)system). Otherwise, you are assuming that energy appears from nothing. In such case *you are stating that the first law is false*. Quite curious affirmation. Just change the textbook. – RodolfoAP Aug 21 '21 at 06:51
  • @RodolfoAP, not it's not two closed systems. In the study of heat engines, you have an infinite heat source and an infinite heat sink. You are confusing the statement of the first and second laws of thermodynamics with what thermodynamics studies. Those laws are about closed systems, but thermodynamics is not the study of closed systems; it is the study of systems with controlled inputs and outputs, with certain parameters fixed and other allowed to change, so long as everything remains in equilibrium. – David Gudeman Aug 21 '21 at 09:13