How can it be solarisation if the negs weren't exposed to light during development?
Exposure before development just makes fogging.
I don't think that's necessarily true, is it? Here's what I'm thinking:
Each grain only cares about whether it got enough light to activate it before it was reduced in development. Pre-exposure, whether you call it "pre-exposure" or "oops", just delivers some of that light more or less uniformly, before development starts; flashing in the developer does the same thing, but only to the grains that haven't been reduced yet.
In shadow areas, there are a lot of these unreduced, because most of the grains aren't activated and aren't getting reduced; so the shadows tend to get lighter (otherwise known as "fogging" if you didn't want it to happen). In the highlights, the earlier in development the exposure happens, the more of them there are; if the exposure happened very early in the development, or (as in the original poster's case) before any development at all, there are a lot of those too, and the highlight values will also be elevated---so they blow out, or if the amount of light is enough, they experience solarisation and turn back into shadows. If I'm not misunderstanding how solarisation and pre-exposure work, I think that's a possible explanation of what happened here.
That said...while I find this chain of events sort of plausible, I can't explain why the people in the background seem to have reversed, but the white chair rail in the foreground and the four specular highlights in the background haven't. On the other hand, I don't know what the original distribution of light really looked like, and maybe the extra exposure was uneven and affected some areas of the negative but not others.
-NT