If you're looking for the scientific community to come to a consensus on what existed before the Big Bang, well, you're going to be disappointed.
As others have alluded to, the Big Bang is not quite the creation story of the universe. Rather, it is the result of an extrapolation that we know we can't continue indefinitely.
This all began back in the early 20th century, when astronomers like Hubble realized the universe seemed to be expanding. At the same time, theorists working on developing general relativity realized the idea of an expanding universe could fit quite nicely in their models.
At this point, it wasn't much of a leap to ask, "Well, if the universe is expanding, what happens if we go back in time? It must have been smaller then." And in fact, those theorists realized the "size" of the universe goes to zero a finite time in the past. At the same time, the temperature and density must necessarily go to infinity.
Now no one really ever thought these zeros and infinities actually described nature. Rather, they were either:
- (1) A sign that the theory was completely off the mark. This was the position of advocates of the steady-state cosmology, one of whom coined the term "Big Bang" just to highlight how ridiculous the notion was.
- (2) A sign that the theory was incomplete. At early enough times, with temperatures and densities unfathomably large, new physics would come into play. The only reason the theory seemed to give infinities was our ignorance of some of the physics. This is comparable to how basic theories predicted an aircraft would experience infinite drag as it approached the speed of sound, and so the sound barrier could only be broken with the inclusion of better physical models in the regime.
The scientific community has reached a consensus that option (2) is right, abandoning option (1).1
How do we know? Well, even if the theory doesn't reliably extend to the beginning of all things, it still makes quantitative predictions based on instances where we can reliably extrapolate back in time. In particular, taking what we know about particle physics, we can ask, "If the universe were originally too hot to form nuclei (it doesn't matter how much hotter), and then it expanded and cooled, what distribution of nuclei would we expect to condense out?" That predicted distribution matches the observed distribution quite well. You can read more about Big Bang nucleosynthesis for instance in this answer to What has been proved about the big bang, and what has not?
But at some point in the extrapolation the consensus ends. If you ask, "What was the universe like in the first $10^{-50}\ \mathrm{s}$?" you will be told, "We have ideas, but we need more of a theoretical framework before we say anything definitive."
If you go even further and ask, "What came before the Big Bang?" of ten cosmologists, you'll get at least a dozen different answers. Some of the more popular (or at least more discussed) theories are:
- This question is not even well defined. In this traditional, pure general relativity viewpoint, asking, "Is this model for the existence of stuff before stuff existed correct?" is pure nonsense. It's like saying "Let ☢ be the greatest positive integer less than 0. Is ☢ a prime number?" Asking about properties of things that don't exist doesn't even make sense, and there is no more "before the Big Bang" than there is a positive integer less than 0 or a real number less than $-\infty$.
- Eternal inflation: Roughly the idea that everything is just expanding, and every once in a while a quantum fluctuation slows down the expansion in a small patch, which becomes the visible universe.
- A cyclic universe: A contracting universe gets very small, undergoing a "Big Crunch," and, once it has extremely high but not mathematically infinite density it "bounces" back in a Big Bang, leaving little trace of the previous incarnation of the universe.
- The ekpyrotic universe: Two string-theory-esque branes collide in some "pre-existing" background, with the result being the birth of the universe as we know it.
So there are many scientific ideas about what came before the Big Bang, but you can't point to a collection of conflicting theories as evidence that a separate theory is wrong.2
1 I feel obliged to point out that supporters of the steady-state theory, while it was still popular, ended up making great contributions to physics all the same. In particular, they pioneered the field of stellar nucleosynthesis in an effort to explain how stars could convert hydrogen to heavy elements. As a result, they also pushed ahead the field of general nuclear physics considerably.
2 If you want to challenge a theory, the following is how you put it to the test. First, the theory must be well defined. It cannot have dangling poetic phrases that don't mean anything beyond sounding "sciency." Second, it must be internally self-consistent, at least within its purported domain of applicability. Most theories rejected by the scientific community right off the bat fail these tests. If the theory passes, the third step is to ask if it can explain observations already explained by the theories it claims to replace. If not, then it is deemed a nice attempt, but not in line with nature. Finally, if the theory passes the first three tests, you ask it to predict something new in such a way that its competitors will disagree with it. If this cannot be done, the theory is not new but rather just a rewording of an old theory (which may or may not be useful to have). If it can be done but the subsequent experiment disproves the new theory, then it is deemed a very nice attempt that just happens to not work with nature. Only if it passes this fourth test is it accepted as a new, working scientific theory, but even then it may be superseded in the future by better theories.