I'm sure everyone working in theoretical physics knows the problem: Many, probably most, of the people who are most important in your life -- family, friends, loved ones -- don't have the background to understand what your research is all about. So what will you do? A strict secret-agent-style "don't talk about work when at home" policy is deeply unsatisfying; after all, you went into academic research rather than into a more lucrative field such as, say, finance, because you are driven by a consuming interest in finding out how the universe works. But any attempt at a conversation about our work tends to end in a muddle of confusion ("What do you mean by a symmetry? What on earth is a group? Why do you have to prove that, isn't it obvious (or clearly false)? What's a simulation? How can a computer solve the equations if you can't, I mean, don't you have to tell it how to solve them? ...?") or worse yet, misunderstandings ("--- studies particles that form groups and have links between them, or something like that!") when talking to people lacking the basic concepts necessary to understand contemporary research, i.e. to virtually all non-scientists. Of course there always is the pop-sci approach (leave out mathematics and every other bit of "technical" detail and just focus on the beauty and wonder of it all), but the people closest to you usually want more than that: they want to be able to ask "How was your day/week?" and get a meaningful answer that does not exclude your research, which is after all what you spend most of your time doing. But if their background is in the arts or humanities, they don't just lack the technical knowledge of, e.g. group theory -- that could be easily remedied; no, most of the time they are actually deeply unfamiliar with mathematical reasoning and the general modes and methods of scientific research.
At Christmastime, there's always the question "what can I give to ---", and what better than to give the gift of greater understanding? For introducing people with a humanities background to the kind of ideas and ways of thinking used in some of the more abstract fields of theoretical physics, I have tried titles from Oxford University Press's Very Short Introduction series of books, such as the one on Particle Physics by Frank Close, the one on Quantum Theory by John Polkinghorne, the one on Cosmology by Peter Coles, or the one on Mathematics by Fields medalist Tim Gowers. I think they do a very good job at giving a non-technical introduction to their respective subjects that goes a good way beyond the usual pop-sci stuff without trying to make experts out of their readers.
For the more down-to-earth kind of physics, the most recent issue of PhysicsWorld (the one with a lattice on the front page), contains a very positive review of Louis Bloomfield's "How Everything Works", which is said to be a general-market version of the author's textbook "How Things Work" for physics courses for non-scientists.