I come from kind of a varied background. Working for more than six years in a library, I had access to a number of books, and some became particularly valuable. Rapid Development by Steve McConnell is one of them. In it, McConnell outlines studies about people and processes, personality, management, and development, all with the emphasis on generating rapid development techniques.
Of interest to me is that McConnell doesn't really go into specific rapid development or agile methods techniques, such as SCRUM or Extreme Programming. Certainly, he discusses some of them, but the book is mostly geared towards eliminating the barriers many companies have to overcome.
He reviews the 36 Classic Mistakes, a part of his book (I believe in chapter 5), that work in concert to prevent successful rapid development efforts. This is one of the most worthwhile reads, and should take at most 30-45 minutes out of your day.
Companies are, of course, all different; some of these don't apply to us at all (for instance, we've been really strong working with source control, and not expecting tools to solve a lot of our development problems). Others afflict us only occasionally - when we have a project that has to come up and down the pipe quickly, we sometimes poorly manage risk. Others are something that we run over regularly.
I mentioned my varied background - I like to understand people. I was a social psychology buff in college, and one of the reasons I am chosen to mediate certain cases is specifically for my ability to understand where people are coming from. We all think of our projects differently - and developers see different things than PMs, than artists. The truth is, these kinds of mistakes hit all of us, and if we can take the time to understand each other a bit more, maybe we can become a more cohesive team.