What Gödels incompleteness theorems stated was that it is impossible for mathematics to be entirely consistent because you still have to make the assumption of why arithmetic is consistent because that is the one thing that it cannot prove. It would be circular reasoning.
But what Gödel also proved in his completeness theorem a while before was that if a statement is logically valid you can only deduct one conclusion from it. For example, if I state "Richard has a driver's license. You can only get a driver's license if you're sixteen.", and everything I said is true, then two people cannot arrive at different conclusions regarding if Richard's sixteen or not (again, the statement has to be built by axioms, you cannot say "well, maybe he got it anyway", because the second premise is that you can only get it if you're sixteen).
So, making the one assumption that arithmetic can describe everything but itself consistently, we can then look at mathematical statements and see if they're consistent. Makes sense? Otherwise, I'll have to refer to Occam's razor wich would say that one assumption is good enough to use mathematics the way we use it today, and my English isn't remotely good enough to go there.