Stephen Hawking says in his latest book The Grand Design that,

Because there is a law such as gravity, the universe can and will create itself from nothing.

Is it not circular logic? I mean, how can gravity exist if there is no universe? And if there is no gravity, how can it be the reason for the creation of universe?

Also, if the universe doesn't exist, how can it create itself? The very sentence doesn't make sense to me. It seems so absurd and illogical that I've never heard such sentences even in philosophy. On what grounds does Stephen Hawking claim this?

  • 5
    $\begingroup$ The quote as stated indeed makes no sense. What Hawking probably meant is that there is no conservation of energy on the spacetime scales. I.e. when you start with something small (such as BB singularity), it's inevitable that you obtain huge universe that arguably contains much more stuff than it originally did (because energy is created as the universe expands, so to speak). Be that as it may, we'd need more context to judge the quote. $\endgroup$ – Marek Jul 31 '11 at 15:59
  • $\begingroup$ Why do You not ask the authors? $\endgroup$ – Georg Jul 31 '11 at 17:08
  • 1
    $\begingroup$ You can't judge a sentence such as that out of its context. Specially if it comes from a pop-science book. $\endgroup$ – becko Jul 31 '11 at 17:17
  • $\begingroup$ I highly recomment this talk by Lawrence Krauss, titled "A Universe From Nothing". I cannot remember if he specifically talks about Hawking's theory, but the idea is the same. $\endgroup$ – Lagerbaer Aug 8 '11 at 21:58

Unfortunately I don't have a copy of "The Grand Design", so can't be 100% sure of the context, however, from the terminology, it sounds like this may be a reference to Hartle and Hawking's no-boundary proposal.

In this scheme, they propose a method for computing what they refer to as the "wavefunction of the universe". This wavefunction uses Feynman's path integral to assign probability amplitudes to three-metrics on a three-surface $\Sigma$ bounding a Euclidean spacetime M. By analytic continuation, the wavefunction can be continued to a function representing a Lorentzian signature spacetime.

This approach is explained in Hawking's publicly available lecture. There he describes an explicit example where $\Sigma$ is a three-sphere and the Euclidean manifold M is a four-ball. "On the other side" of the bounding three-sphere $\Sigma$ is Lorentzian de Sitter space. This model is proposed as a model for a spontaneously created de Sitter universe, and he makes the statement

Unlike the black hole pair creation, one couldn't say that the de Sitter universe was created out of field energy in a preexisting space. Instead, it would quite literally be created out of nothing: not just out of the vacuum, but out of absolutely nothing at all, because there is nothing outside the universe.

  • 1
    $\begingroup$ Interesting. Could you elaborate a bit on the creation part though? How is the creation explained in this model? $\endgroup$ – Marek Aug 1 '11 at 13:24
  • 1
    $\begingroup$ @Marek: I'm not sure that "creation" is an accurate way to phrase it. When I think of "creation" I think of going back in time - beyond a certain point the universe didn't exist, then a bit later it did. I don't think the Hartle Hawking picture is like that - it's more like you can go so far in Lorentzian time, then after that the appropriate description is the Euclidean one, for which there is no past boundary. It's because there isn't a boundary that I would be hesitant to use the term "creation". (Although Hawking did, so maybe I'm misunderstanding something !) $\endgroup$ – twistor59 Aug 1 '11 at 13:53
  • 1
    $\begingroup$ Yes, that's my view as well. That's why I don't quite get relation of this answer to the question. $\endgroup$ – Marek Aug 1 '11 at 14:07
  • $\begingroup$ I understand this is old, but dies anyone have an updated working link to this article? I'm interested in self-containment. $\endgroup$ – Goodies Jan 6 '15 at 20:07
  • 1
    $\begingroup$ Nice explanation, though I thought that Hawking - when mentioning this in popular lectures or books - sometimes refers to the 'zero-energy hypothesis', the idea is that it would perhaps be possible to define some form of negative potential energy with gravity, in such a way that the total energy content of the universe would add up to exactly zero. I wouldn't necessarily advocate this myself, but I thought I'll drop it here in case anyone is interested $\endgroup$ – JgL Jun 9 '17 at 19:19

At face value, I interpret the statement as follows:

Quantum mechanical fluctuations may generate from the vacuum ( nothing) enough energy for a universe. Because gravity exists for any ensemble that has energy, then the big bang will go on its usual development in time.

One has to presuppose that the mathematical formulations exist,( a mathematical theory of everything including quantum mechanics and gravity), irrespective of the existence of matter/energy. A kind of platonic ideal.

  • $\begingroup$ Seems to me that one has to presuppose more than that the mathematical formulations exist... one has to presuppose that gravity (and QM etc.) have the power to cause events. Don't you think? And doesn't "irrespective of the existence of matter/energy" beg the whole question? $\endgroup$ – LarsH Sep 7 '14 at 10:41
  • $\begingroup$ @LarsHave you heard of the platonic ideal? en.wikipedia.org/wiki/Theory_of_Forms $\endgroup$ – anna v Sep 7 '14 at 10:55
  • $\begingroup$ Yes... but I'm still trying to think how that applies here (and wondering if that's really what Hawking had in mind). $\endgroup$ – LarsH Sep 9 '14 at 2:36
  • $\begingroup$ @LarsH "one has to presuppose that gravity (and QM etc) have the power to cause events" is exactly what "form" precedes "real world" is about. The mathematics are "form", a format that will manifest reality. I do not know whether he intends to say that, but that is what the quote you give says. $\endgroup$ – anna v Sep 9 '14 at 3:51
  • $\begingroup$ Could you elaborate more on what you mean by "quantum fluctuations generating enough energy " ? It's slightly misleading since it seems like energy conservation is being violated (or is that what you intend to say ?) $\endgroup$ – alex Jun 10 '17 at 18:09

The flaw in the argument is the assumption of causality, and that it's linear. Nothing doesn't exist, which makes saying a given entity is nothing empty.

  • $\begingroup$ The assumption of linear causality is what has made science possible. Is Hawking then abandoning science? $\endgroup$ – LarsH Sep 7 '14 at 10:43

If the input to cosmology is trivial in the sense that the initial conditions have to be trivial, then creation itself was trivial, and needs no explanation. It's the output which needs to be explained, not the input.

There's no need for even the laws of physics to be included in the input because eternal inflation and the multiverse of the string theory landscape leads to an ensemble of all possible laws.


Stephen Hawkings is a physicist and he needs to put any idea of God outside of the domain of physics. (a)
Georges LeMaitre was (WP Hubble-Law) a belgian priest and the BBT leaves intact the divine intervention as a justification to the act of the creation of the universe/matter, and it appraises to the Catholic Church and other religious positions.
If you beleive in ONE single BB then you are a priviledged observer of something that begs the question: And before? and why?
(gravity in the sentence is not about gravity but only: there is a law...)
The sentence:

the universe can and will create itself from nothing.

Is a circular reasoning as you say, even worst: create itself (like god!); from nothing? (a free lunch!). From 'nothing' we can not get 'something'.

I think that Spinosa's view is more adequate to a physicist:
The Universe was, is and will be ALL with no start nor end. The event of matter creation, we use to call it BB, is a repeatable event. This way I do not need to call for any entity exterior to the universe.

(a) - if God exists, or not, is a different question, and is outside of the domain of this site.
In Physics we can not invoke any transcendental entity to be 'a cause'.

  • 3
    $\begingroup$ This depends highly on how people define god, and what properties they attribute to it. $\endgroup$ – Lagerbaer Aug 4 '11 at 22:08
  • 3
    $\begingroup$ Note that Hawkins does not argue for or against god; only whether or not a creator seems necessary or not. In the same way, we still haven't 100% proof that Zeus does not exist, but we know that we don't need Zeus to explain thunder and lightning. $\endgroup$ – Lagerbaer Aug 5 '11 at 4:22
  • $\begingroup$ @Lagerbaer Please reread my Answer to check that you said the same as I did. If the downvote is yours I do not understand the why. $\endgroup$ – Helder Velez Aug 8 '11 at 17:31
  • 1
    $\begingroup$ The downvote is not mine. $\endgroup$ – Lagerbaer Aug 8 '11 at 21:55

protected by Community Aug 7 '11 at 2:42

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.