People can generally not hear ultrasonic tones, but they can dectect pulse trains at much higher than ultrasonic frequencies. They sound like a chirp.
In the case of the Dukane pinger, you don't even need to hold it against the bone behind your ear to hear the once per second tick; it sound like a fairly loud clock. Nothing special to that, the square wave pulse is obvious, and I suspect it is the end diaphragm of the device popping.
What I find interesting is that once you do put it against the bone behind your ear, and if you listen really carefully in a quiet room, you will start hearing the actual tone that comprises the pulse. Imagine the highest pitch squeak you can, and then picture it getting higher in pitch still. That's what it sound like. It is a very odd experience.
This doesn't really surprise me, though. The human ear is a very efficient acoustic filter, optimized for the frequencies we're supposed to hear. It does a pretty decent job of reducing the 'out of band' signals that we're not really intending to listen for (the text book "Fundamentals of Acoustics" covers the math pretty well; my copy is at the office so I'll post the authors name tomorrow night). But when you bypass the filter and put the input directly into the detector, finding that the little cilia in there vibrate at a much higher frequency than expected isn't shocking to me.
I have a few transducers that will operate out well past 50 kHz. I'll have to try a few sine waves (no impulse to confuse things) and see what I can hear. My cats may not like the experiment (the always sit outside the door to the workshop and meow when I play with ultrasonics), but I think I can confuse them for a few minutes for science.