Anyway, it's also possible to analyse the Wien bridge oscillator like I did the filters earlier, by considering capacitors to be generalized resistors, and by considering the part of the circuit below the output to be a filter. But at the end you need to find both the gain and the phase shift. The oscillator condition is for a phase shift of zero degrees (so what the amplifier adds back, without inversion, is in phase with what you start with -- the op amp itself has no phase shift because there are only resistors). Then you find that the filter gain is 1/3, so the op amp gain needs to be 3 (=1 + y/x -> y/x = 2) for stability. I can post the math later.