In a 1931 letter to his sister, the celebrated paleontologist George Gaylord Simpson was pondering the ultimate scientific question, how did the universe begin: "Call that great Unknowable by any name you wish, call it X, or Yahweh, or God, or say that God created it. Applying the letters "g", "o", and "d" to it or what created it is no explanation and no consolation. It is a common failing, even more among scientists than among laymen, to think that naming a thing explains it, or that we know a thing because we can put a name to it. But to say that God created the universe means nothing whatever."
Faced with the (present) mystery of the Big Bang, the empirical naturalist will say "I don't know." Perhaps an explanation will come along, perhaps not, but to say "God did it" adds nothing to our understanding. "If a sign is useless, it is meaningless; that is the point of Ockham's razor," said Wittgenstein.
To name a perceptible thing has some advantage; it makes it possible to talk about it. There can be no theory of the electron, for example, until we have a word for the electron. But naming is not understanding. Before we say we understand a thing, we must weave it into the web of concepts that constitutes a theory. Only when the concept "electron" is enmeshed in a matrix of other ideas -- atoms, fields, valency, molecular bonds, etc. -- by taut, quantitative connections, do we have confidence that we know what an electron is.
Is there a circularity in scientific explanation? Of course. Every explanatory system refers back upon itself. It is the timbre of the web and the way the web makes empirical verification possible that give us confidence that we are doing something right. To say that "God" caused the Big Bang predicts absolutely nothing about what we should see when we turn our telescopes to the most distant universe. The thread of meaning that connects "God" to the earliest universe is infinitely slack.