Steering, as in yaw, you pretty much will always need some type of vertical surface to accomplish. You can also use split ailerons or clamshell rudders to create differential drag on either side of a pure flying wing - it's something I've experimented with in the past, and it works quite nicely if executed properly. However this defeats the (i assume) purpose of this which is to eliminate control surfaces from the wing.
You can turn with just roll and pitch, known as banking and yanking. It creates inherently uncoordinated turns most of the time, but they are turns. Now, the problem here - is the reason control surfaces are generally small compared to the total size of the wing is to allow them more movement while not creating an excessive amount of authority on any control axis, thereby allowing more precise manipulation of the airplane's attitude. Sure, you can control an airplane by altering the angle of incidence of the wing, though the problem is the travel of the wings would have to be incredibly small.
Something kind of like this actually has been done before. In the 1930s there was this weird airplane called the Granger Archaeopteryx, which used the entire outward portion of the wing as an aileron. The entire wingtip was rotated for pitch and roll control. I read somewhere that this apparently made the plane more resistant to spinning though I'm really not sure how.
