Of course this is one of the oldest philosophical questions in science but still one of the most mysterious. For most of Western history the cannonical answer has been some version of Platonism, some variation on the esentially Pythagorean idea that the matherial universe has been formed according to a set of transcendent and a priori mathematical relations or laws. These relations/laws Pythagaoras himself called the divine armonia of the cosmos, and have often been referred to since as the "cosmic harmonies" or the "music of the spheres". For Pythagoras numbers were actually gods, and the quest for mathematical relations in nature was a quest for the divine archetypes by which he believed that matter had literally been in-formed. Throughout the age of science, and even today, most physicists seem to be Platonists. Many are even Pythagoreans, implicitly (if not always with much concious reflection) making an association between the mathematical laws of nature and a transcendent being. The common association today of a "theory of everything" with "the mind of God" is simply the latest efflourescence of a two and a half millenia-old tradition which has always viewed physics as a quasi-religious activity.
Can we get beyond Platonism in our understanding of nature's undeniable propensity to realize extraordinarily sophisticated mathematical relations? Although I began my own life in science as a Platonist I have come to believe that this philosophical position is insupportable. It is not a rationally justifiable position at all, but simply a faith. Which is fine if one is prepaared to admit as much, something few physicists seem willing to do. To believe in an a priori set of laws (perhaps even a single law) by which physical matter had to be informed seems to me just a disguised version of deism — an outgrowth of Judeo-Christianity wrapped up in scientific language. I believe we should do better than this, that we should articulate (and need to articulate) a post-Platonist understanding of the so-called "laws of nature." It is a far from easy task, but not an impossible one. Just as mathematican Brian Rotman has put forward a post-Platonist account of mathematics we need to achieve a similar move for physics and our mathematical description of the world itself.