The analogy of "if it was a spoken word, it would sound as if it was getting played backwards" was helpful, but I'm still struggling to understand what's happen

The analogy of "if it was a spoken word, it would sound as if it was getting played backwards" was helpful, but I'm still struggling to understand what's happening here. Thinking of the audio analogy, I could imagine implementing that as a buffer you record into, and after some amount of time, you play back the buffer in reverse. Is that what's going on here?