This is an issue that is killing me. I know that…:

Let K=“I know”

- K(p) & K(q)
- K(p & q)

…is an invalid form, because K is referentially opaque (if you don’t know what I’m talking about, this will help http://www.artsci.wustl.edu/~philos/MindDict/referentialopacity.html), but is this next example a valid form?:

1’) K(p & q)

2’) K(p) & K(q)

My first gut instinct would be to say no, but I can’t seem to find a counter example to it, and its driving me insane.

If anybody has any idea what is going on here, I would like to know. It seems to me that I should be able to preserve truth backwards and forwards, but maybe thats just an illusion I need to drop.