Did the Big Bang violate the second law of thermodynamics?


Creationists often claim that (i) Big Bang theory posits a high degree of order in the initial state of the universe, but that (ii) the second law of thermodynamics forbids this, and that the Big Bang therefore could not have happened naturally.

The short response

  1. If the universe ever was part of a larger system, then it could have purchased low entropy at the expense of increased entropy in the larger system.
  2. If time itself began with the intial state of the universe, then low initial entropy is consistent with the second law of thermodynamics.
  3. The expansion of spacetime can cause the maximum possible entropy of the universe to grow, creating room for local order even if the initial state state of the universe had as high entropy as was possible at that time.

The longer response

The second law of thermodynamics states that the entropy of a closed system never decreases over time. This definition makes evident a number of problems with the creationist line of reasoning.

I. Entropy vs. disorder

As Lorenzo Benito has emphasized to me, it is important to be clear that the second law of thermodynamics deals with entropy rather than with “disorder.” Although entropy often is described as a measure of disorder, only a very technical kind of disorder matches entropy; there is no rigid thermodynamic correlation between entropy and intuitive notions of disorder such as those invoked by conversational examples like shuffled decks of cards and messy rooms. For instance, entropy increases when one shuffles an ordered deck of cards not because of the randomization of the card faces but because of the heat energy released by one’s muscles while shuffling. While there are reasons to expect shuffling to randomize cards, those reasons are not thermodynamic ones. John Pieper’s talk.origins article Entropy, Disorder, and Life discusses in detail the limitations of the relationship between entropy and disorder.

It is not difficult to recast the creationist argument in terms of low and high entropy instead of order and disorder, but the modified argument still runs into problems as described ahead.

II. Entropy can decrease in open systems

The creationist argument ignores the possibility that the universe is an open system. If the universe is, or ever was, part of a larger system (as hypothesized, for instance, by Smolin 1999, in which universes reproduce), it could have purchased its initial low entropy at the expense of increased entropy in the remainder of the larger system. This is something that can in principle continue forever. For instance, in Steinhardt and Turok’s cyclic theory:

The total entropy of the universe as a whole increases steadily from bounce to bounce, as demanded by the second law of thermodynamics. However, the entropy from the previous cycle is spread to regions beyond the horizon during the period of dark energy domination and never seen again. So, as far as any local observer is concerned, the entropy density is driven to zero before each bang and the universe appears to begin afresh after each bang. (Steinhardt and Turok nd.)

Some creationists insist that only open systems with some kind of sophisticated conversion mechanism in place can undergo a decrease in entropy. However, that claim (1) is neither stipulated nor entailed by the second law of thermodynamics and (2) is in fact known to be false from examples like crystallization and endothermic reactions. It is important to be attentive when presenting these examples to creationists because many respond by playing a shell game, replying that crystallization and endothermic reactions do not produce “complex information.” But remember that the initial claim concerns what the second law of thermodynamics permits or forbids: the second law of thermodynamics says absolutely nothing about complex information.

III. What if the universe is a closed system?

If the universe is a closed system, a violation of the second law of thermodynamics would require that there be two distinct states of the universe, U and U*, with U preceding U*, for which U is of higher entropy than U*. But if we go by the equations of general relativity alone, then the origin of the universe was the origin of time itself), which means there could have not have been any state prior to the initial state of the universe, much less one with higher entropy than a subsequent state. This goes likewise for scenarios employing quantum gravity in which there was a first moment of time: for any cosmological model in which there is a first moment in time, the initial state of the universe can be of as low entropy as one wishes without violating the second law of thermodynamics.

Theories of quantum gravity that allow the universe to continue existing infinitely into the past are a different matter. It is unclear to me whether there is any way to reconcile them with the second law of thermodynamics, or whether they would require abrogation of the second law of thermodynamics.

Theories of quantum gravity that have time breaking down in the early universe are also difficult to assess, but since the second law of thermodynamics makes essential reference to time, it would seem that the second law of thermodynamics could not apply to such cosmological models at all.

IV. Expansion of a closed universe

Physicist Victor J. Stenger points out (2007: 117-121) that in an expanding universe, it actually is possible for the universe to start out in a state of maximal entropy and still increase in entropy over time because the expansion of spacetime also increases the maximum allowable entropy. In fact, the maximum allowable entropy increases faster than the actual entropy, meaning there is room for local increases in order.


Smolin L. 1999. The Life of the Cosmos. Oxford: Oxford University Press.

Steinhard P and Turok N. nd. Endless universe: frequently asked questions. https://www.physics.princeton.edu/~steinh/endlessuniverse/askauthors.html

Stenger VJ. 2007. God: the Failed Hypothesis. Amherst, NY: Prometheus.