A few questions on low-level engineering of networks

Hi guys, I'm a CS student and finally got to take an "upper level" networking class, and am loving it, but there's one thing I still can't quite get down clearly. I have a great book but it doesn't detail these doubts as clearly as I'd like. I have a few questions I'm hoping can be cleared up:

1. How does circuit switching/packet switching actually transmit bits? Say it is running on a wired connection, how does the cable actually transmit binary from one host to another? What clearly differentiates the two?
2. What does "switching" actually mean? I don't understand it in this context, does it just mean the transfer of information?
3. What is the reason we prefer packets over a steady stream of bits? Is it because packets include header information that allows the end destination to more efficiently/quickly build back up the entire data, instead of piecing it together without context? (As well as more easily identify data loss)
4. In the context of 3., how does an end host know when to start/allocate space for a packet and then end (wrap up packet and move to next) - are these just signals that are asserted?

Thank you so much, and apologies for the low-level questions, but I'd feel much better about my knowledge if I understood these things, they're really bugging me. Any suggested reading is also highly appreciated!