Is it possible for some being to violate the laws of physics? My belief is that it isn't possible for any being, no matter how powerful, to violate the laws of physics. Because, simply by definition, the laws of physics apply everywhere and at all times. But what have philosophers written about this matter?

  • If the "being" is not "natural", maybe. Commented May 29 at 14:01
  • 1
    The laws of physics don't apply everywhere and and at all times, especially not by definition. There are places (e.g. black holes) and times (e.g. big bang) where there are singularities which mean that the laws of physics break down and we don't know what actually happens there. We also don't know the laws of physics (at least for certain) so isn't this just defining anything that happens as being the result of some laws of physics of which we only have uncertain knowledge? Commented May 29 at 14:33
  • Alternatively we may live in a computer simulation and the programmer has left a back-door so the program allows her avatar to break the rules. The definition of what is meant by the laws of physics is unclear here. Commented May 29 at 14:34

3 Answers 3


One of the "default" (but far from sufficient) characteristics of a "law of physics" is a universal quantification (modulo if-then), e.g., "Every occurrence of Y's is the effect of X's," or, "Every time X happens, Y happens as a result." To "break" a law of physics is to provide a counterexample, then, a case of some X not causing some Y, or of some Y happening to be found without its usually corollary X. So there is a trivial sense in which a law of physics, as a true proposition, could not be broken or else it wouldn't be true in the first place.

However, let us hedge our bets and speak of probabilistic laws, too. On one level, to say that something happened in defiance of the stronger probabilities, but in compliance with the totality of the probabilities, is not what we would generally mean by "breaking a law"; on the other hand, the individual event-tokens that we would be tempted to see as "law-breaking" could nevertheless conform well to the definition, so we might replace the concept of breaking-a-law-of-physics with defying-the-main-probabilities managed by the laws.

And in yet another context, this all testifies against equating miracles, or supernatural powers, with "law-breaking." For example, for resurrecting someone to "break the laws of physics" would require there to be some law like, "No human ever comes back from the dead." It would be more appropriate to say something like, "A miracle is such as when a physical Y is caused by a non-physical X," which can be an independently implausible claim, but need not be taken for something so drastic as the "violation" of a law.

A more concrete example: consider the matter of the speed-of-light. Perhaps the most naive way to put it is:

  1. Nothing can travel faster than light.h

But this isn't quite what the actual full theory says. The actuality is more like:

  1. Nothing can travel faster than light except tachyons.

Tachyons might not happen to exist, but on some level, they do seem possible. Unfortunately, on the other hand, (2) is ambiguous between the way it was just stated and the tautology, "Nothing can travel faster than light except things that can (indeed must) travel faster than light." The nontriviality, such as it is, comes from the prefix "tachy-", which reflects the fact that particles that travel faster than light would go the opposite direction in time relative to particles traveling at or below light speed. So the full "law of physics" would perhaps best be stated as:

  1. Every particle that travels faster than light also travels in the opposite direction through time.

By analogy, we might qualify a claim about someone's resurrection in terms of a law like:

  1. Every dead person who comes back to life is brought back by divine power (if this ever occurs).

... rather than:

  1. No one can come back to life.

hAlternatively, "Nothing that starts out moving slower than light can ever reach light speed by a gradual increase without acquiring infinite mass and thus 'self-destructing'." But then we could go on to qualify this by saying, "Nothing can increase to light speed, gain infinite mass, and survive this process unless abstract force X preserves it," where X is some convoluted speculative phenomenon that would preserve some relevant symmetry of X despite the acceleration to light speed and acquisition of infinite mass. Again, regardless of whether such a thing ever does happen, there is some level on which it does seem possible.

And to make the question even more complicated: what of the partial difference between, "A moves from x to y at speed z," and, "A changes location from x to y, with no issue of speed in play" (i.e. A just "teleports" from x to y)? There is a non-technical sense in which teleportation is "faster" than motion, e.g. it would be faster than traveling at light speed even were that 4,815,162,342 kilometers a second rather than ~300,000 km/s in the sense that it would "take less time" to teleport across 1 lightyear's distance than it would take for light to move that distance. But the possibility of a particle's wave-function collapsing in such a way that the particle-wave's initial state was more in the vicinity of x but the post-collapse state was localized closer to y (for a standard example) is not a violation of a limitation on motion. And so perhaps we might distinguish between forms of types of "miracles," like A-resurrection and B-resurrection, and say that the proposition of someone's A-resurrection is inconsistent with the physics of A-processes, whereas someone's B-resurrection has the same result as the A-process would except by a different "vehicle" (as when being-in-position-y is variously taken to be the result of motion through, vs. teleportation across, space).

  • 1
    +1 The OP should keep in mind laws of the universe are linguistic descriptions based in part on inductive reasoning, not real things like apples and oranges, and that the sciences actively look for violations of laws to ensure it is descriptions are accurate. Inductive reasoning has a well-known "problem".
    – J D
    Commented May 29 at 16:16
  • 1
    @JD "not apples and oranges"... c'mon... Newton lived in England... expecting "apples AND oranges" be involved is a bit much dontcha think? At least he included the apples. Commented May 30 at 5:45

A phenomenon which violates the laws of physics is named a “miracle”. Hence you ask about the possibility of miracles.

  1. History of science knows several discoveries which show the limit of existing scientific theories and scientific conceptions. But these contradictions should not be mixed-up with miracles. Each theory has a range where it applies. Within this range the theory makes a general claim, excluding contradicting exceptions.
  2. On the other hand, history knows many reports of miracles. Each religion reports miraculous acts of its god or his devotees to demonstrate the power of the god. But these reports are not reliable, because they do not satisfy the requirements for a historical source.

To sum up: Unbiased reliable reports of miracles do not exist. Each miracle would question one or more physical laws from the corresponding domain.


Kristian Berry's answer is good, but I think it may help to reduce the problem to its most boring form: definitions, identity, and noncontradiction.

Laws of physics has two meanings: the way things always behave; and the human attempts to guess the way things have behaved so far and predict the way they will behave in the future.

This question is therefore two different easy questions, one for each definition.

1: Can a thing behave other than the way it always behaves? No, this is a contradiction.

2: Are humans fallible? Yes.

If one wishes to define the laws of physics to only describe the way a part of reality behaves, then 1 becomes:

1a: is there a part of reality which behaves the way that part of reality behaves? Yes, this is an identity.

And if we don't know whether or not there is such a part, then 1 defaults to the question of whether reality has such parts:

1b: Is there is a part of reality which behaves the way that it would behave if it existed? If there is such a part, yes. Else no.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .