This is an interesting question. I would start off by saying that question that are based off a either / or approach have some problems. Often times, the problem is a false binary, as in this situation. In other words, why does it have to be one or the other? Why can it not be both?
At times the identification of a problem comes first, but at other times identification of a problem only comes after more is understood. In other words, it is research that shows the complexities of the problem that are in view. And this leads to further understanding of problems and further question.
In addition to this point, we need to also realize that each discipline differs as well. Some fields are so complex that problems usually come much later as people begin to understand the complexities. One example would be anything field that deals with people.