"Splitting" of natural frequency
If you are talking about a split of frequency measured experimentally, this can happen. There could be several reasons, such as feedback to the control input at resonance. More likely, there may be an interface fastener bolt that has lost some torque and backed off, causing the split, with the average being the theoretical frequency. I ran a test once with a mode at 140 Hz; the next time it spilt into three peaks averaging 140 Hz; we tightened all the bolts and it went back to one fequency.
Again, this is one of several possibilities; tests always bring out true colors compared to theory!