I'm not an expert, but I think I know what the deal is.

First off, the voltage of a single line has no meaning except relative to the voltage of another line. (Just like, say, "in motion" or "at rest" have no meaning except relative to a frame of reference. "At rest" relative to, say, the earth or the sun or the galactic center or car interior or whatever.) If you step on one rail of an electric rail line, nothing happens. You don't even feel anything. If you step on the other, similarly nothing. If you bridge both, you're gonna have a bad time</southpark skier voice>.

"Ground" isn't zero volts in any objective sense. And Vcc isn't 5 volts (or 3.3 volts or whatever) in isolation.

And in fact, it's only convention that ground is arbitrarily labeled "zero" and Vcc is 5 volts rather than Vcc being labeled zero and ground being labeled -5 volts. (Ok, it's a little more than just convention. The latter would be awkward in practice. But it would be consistent, and you could do all your circuit design that way, just like if you labeled the white keys on a piano "Z", "Y", "X"... instead of "A", "B", "C"... You'd just have to think differently about it to make it work.)

So, I think what's going on here, and we'll pretend "Vcc" is 5 volts (relative to ground) just for the sake of this example, is that basically:

• The voltage differential between ground and Vcc is 5 volts.
• The voltage differential between Vee and ground (and the order in which I list "Vee" and "ground" here is important) is 5 volts. (If I listed "ground" and "Vee" the other way around, it'd be -5 volts.)
• The voltage differential between Vee and Vcc would be 10 volts.

So, for instance, if the device has some components that operate at 5 volts and some that operate on 10 volts (I dunno, let's say the CPU requires 5 volts, but the... I dunno, optical drive motor?... requires 10 volts), you can get 5 volts by using Vcc as the source and ground as the sink, and you could get 10 volts by using Vcc as the source and Vee as the sink.

That all make sense?

And to one of your specific questions:

Is this a figure of speech or can voltage actually have a negative value?

So, what qualifies as "negative" and what qualifies as "positive" is just convention. Just like what qualifies as the "north" end of a magnet and what's "south" is convention. (I guess "north" and "south" refer to which direction a magnet points when used as a compass, but then again which end of the Earth is "north" and which end is "south" is just linguistic convention as well.)

In the same way, what we label "positive" voltage and what we label "negative" is convention. However, we do have an established convention with voltage. (And with poles of a magnet.) Just because we "made up" what qualifies as "positive" or "negative" doesn't mean there's any disagreement on the topic. So unless you purposefully switch your language just to be contrary, you're not going to get funny looks or requests for clarification using the terms "positive" and "negative" for electrode voltages.

However, if line A has a lower voltage than B (and what's "lower" or "higher" is convention) we can say that the differential between B and A (again, order is important) is negative whereas the differential between A and B is positive.

The analogy between voltage and pressure only goes so far, but in this case I think it's helpful. If tank B has higher pressure than tank A, you can say the pressure differential between B and A is negative. Similarly, you could say that the air pressure differential between sea level and the peak of Mt. Everest is negative. (Though, with pressure, there is such a thing as "absolute zero". "Perfect vacuum" is technically a meaningful concept. I don't think there's any such thing as "absolute zero" for voltage. Only "differences" between voltages on different electrodes.)

It means you have the equation backwards.