Ben -- thanks for your interest and your kind words.
You're right that Shannon's approach has been hugely influential. To become a useful concept in the physics of the last century, communication had to be reduced to the transmission of quantifiable data. What's unfortunate is that "the semantic aspects of communication" -- i.e. what makes information meaningful -- has rarely been connected with the physical issue of measurement. That involves a different and more complicated "engineering problem" -- how to set up the physical contexts that let particular kinds of information be determined. We tend to think of "meaning" as something we humans project onto the world.
Conrad