Note that Craig is not saying that everything must have a cause, which is a frequent misinterpretation. Only something with a beginning is supposed to require a cause. However, he gives no reason for (1) other than a kind of "metaphysical intuition."
While (1) can be challenged on a number of fronts, let me just mention one rebuttal that has been made from physics. Quantum electrodynamics is a fifty-year-old theory of the interactions of electrons and photons that has made successful predictions to accuracies as great as twelve significant figures. Fundamental to that theory is the spontaneous appearance of electron-positron (anti-electron) pairs for brief periods of time, literally out of "nothing." Thus we have a counter example to statement (1), something that begins without cause.
However, one might still argue that the quantum process, though random, is still a
causal one. Rather than engage in a semantic dispute on this, let me move to
statement (2) of the kaläm syllogism and attempt to show that the universe did not
necessarily have a beginning. This will also serve to rebut another theist claim:
Ross also uses the kaläm argument to counter the common atheist taunt: "Who created God?" He claims that God is not confined to a "half dimension" of time, and so need not have been created. I take this to mean that if I can demonstrate the universe is not necessarily confined to a half dimension, then Ross, Craig, and other theists who use the kaläm argument will be forced to admit that the universe was not necessarily created. (Of course they won't).
For this purpose, it should be adequate for me to provide a scenario in which the universe occupies both halves of the time axis. I do not feel compelled to prove that this scenario is true, just show how this is possible within the framework of existing knowledge.
My scenario is provided by inflationary big bang cosmology. Ross, Craig, and I agree that the big bang is strongly supported by astronomical observations. Inflation remains less firmly established, but remains the only current theory that successfully explains a wide range of observations. Furthermore, the model is falsifiable, and so maintains good scientific credentials. Indeed, with the 1992 COBE observation of a 1/100,000 fluctuation in the temperature of the cosmic microwave background, inflation passed at least one risky falsification test.
Suppose the universe was at some point in time completely empty of matter, radiation, or energy of any type. It was about as nothing as nothing can be, a void. Physicists can still describe the void in terms of general relativity. It is completely flat geometrically, with space and time axes that run from minus infinity to plus infinity. Anything else and matter, radiation, or spacetime curvature would have to exist and this universe would no longer be a void.
In the absence of matter and radiation, Einstein's equations of general relativity yield the de Sitter solution, which simply expresses the curvature of space as proportional to the cosmological constant. When the universe is flat, this term is zero and the equation then reads: 0 = 0. This denotes the void.
This is the way things would have stayed were it not for quantum mechanics, which we can also apply to an empty void. The uncertainty principle allows for the spontaneous, uncaused appearance of energy in a small region of space without violating energy conservation . If that energy appears as matter (that is, rest energy) or radiation (kinetic energy of massless particles like photons), then it will have to disappear in a short time interval to maintain energy conservation. This can be expected to happen randomly throughout the spacetime void, with no significant permanent result.
However, another possibility exists that can lead to a quite significant and permanent result. The fluctuation energy can appear instead as spacetime curvature within this tiny region, which is called a "bubble of false vacuum." This bubble still contains no matter or radiation, but is no longer a "true vacuum" because of the curvature, as expressed by a non-zero cosmological constant. According to the de Sitter solution, the bubble will expand exponentially in what is called inflation.
The energy density is constant for a brief interval of time. As the volume of the bubble increases exponentially during that interval, the energy contained within also increases exponentially. Although the first law of thermodynamics may seem to be violated, it is not. The pressure is negative and the bubble does work on itself as it expands. By the time it has inflated to the size of a proton, in perhaps 10 42 second, the bubble contains sufficient energy to produce all the matter in the visible universe today. Frictional processes (this is all in the equations--see Stenger 1990) bring inflation to a halt, particle production begins, and the familiar Hubble expansion of the big bang takes over.
We can label as t = 0 the time at which the initial quantum fluctuation takes place. The expansion then proceeds on the positive side of the t-axis, as defined by the increasing entropy on that side. As first suggested by Boltzmann a century ago, the direction of time is by definition the direction in which the entropy of the bubble universe increases.
Now, what about the negative side of the t-axis, the other half dimension? If we look at Einstein's equations, nothing forbids an expansion in that direction as well. Physicists usually simply ignore that solution because most share Ross's prejudice, expressed above, that time "proceeds only and always forward." But the equations of classical or quantum physics, including those of general relativity, make no fundamental distinction between the two time directions. Where that distinction appears, it is put in by hand as a boundary condition.
However, a completely time-symmetric solution of Einstein's equations for the vacuum will give exponential inflation on both sides of the time axis, proceeding away from t = 0 where the initial quantum fluctuation was located. This implies the existence of another part of our universe, separated from our present part along the time axis. From our point of view, that part is in our deep past, exponentially deflating to the void prior to the quantum fluctuation that then grew to our current universe. However, from the point of view of an observer in the universe at that time, their future is into our past--the direction of increasing entropy on that side of the axis. They would experience a universe expanding into their future, just as we experience one expanding into our future.
Would these different parts of the universe be identical, kind of mirror images of each other? Not unless physics is completely deterministic, which we do not believe to be the case. The two parts would more likely be two very different worlds, each expanding in its own merry way, filled with all the other random events that lead to the evolution of galaxies, stars, and perhaps some totally different kind of life.
This scenario also serves to explain why we experience such a large asymmetry in time, while our basic equations do not exhibit an asymmetry at all. Fundamentally, the universe as a whole is time-symmetric, running all the way from minus eternity to plus eternity with no preferred direction, no "arrow" of time. Indeed, the whole notion of beginning is meaningless in a time-symmetric universe. And, without a beginning, the kaläm cosmological argument for a creator fails because of the failure of step (2) in Craig's syllogism.
I have described a scenario for an infinite, eternal, and symmetric universe that had
no beginning. The quantum fluctuation occurs at one particular spatial point in an
infinite void. Obviously it could have happened elsewhere in this void as well. This
multiple universe scenario is exactly what is suggested by the chaotic inflationary
model of Andre Linde. While multiple universes are not required to deflate the
kaläm argument, they can be used to provide a scenario by which the so-called
anthropic coincidences may have arisen naturally. Again, this scenario cannot be
proven, just presented as a possibility that provides a non-supernatural alternative
to the theistic creation. For more discussion and further references, see Stenger
Finally, by showing that the universe did not necessarily have a beginning, we can counter another common theist line of argument used whenever the claim is made that a spontaneous "creation" violates no known physics. The theist will say, "Where did physics come from?" If their imagined God did not have to come from something, because she had no beginning, then neither did physics.
Ross, Hugh 1995. The Creator and the Cosmos: How the Greatest Scientific Discoveries of the Century Reveal God (Colorado Springs: NavPress).
Stenger, Victor J. 1990. "The Universe: The Ultimate Free Lunch." European Journal of Physics 11, 236-243.
Stenger, Victor J. 1999. "Natural Explanations for the Anthropic Coincidences."
Submitted to Philo. Preprint can be found at