Lakos, John. Large-Scale C++ Software Design. Boston, MA: Addison-Wesley, 1996. Lakos argues that a system's "physical design"—its hierarchy of files, directories, and libraries—significantly affects a development team's ability to build software. If you don't pay attention to the physical design, build times will become long enough to undermine frequent integration. Lakos's discussion focuses on C++, but the insights related to "physical design" apply just as much to projects in other languages.

Myers, Glenford J. The Art of Software Testing. New York, NY: John Wiley & Sons, 1979. This classic testing book discusses integration as a testing activity.

McConnell, Steve. Rapid Development. Redmond, WA: Microsoft Press, 1996. Chapter 7, "Lifecycle Planning," goes into much detail about the tradeoffs involved with moreflexible and less-flexible life-cycle models. Chapters 20, 21, 35, and 36 discuss specific life-cycle models that support various degrees of incrementalism. Chapter 19 describes "designing for change," a key activity needed to support iterative and incremental development models.

Boehm, Barry W. "A Spiral Model of Software Development and Enhancement." Computer, May 1988: 61–72. In this paper, Boehm describes his "spiral model" of software development. He presents the model as an approach to managing risk in a software-development project, so the paper is about development generally rather than about integration specifically. Boehm is one of the world's foremost experts on the big-picture issues of software development, and the clarity of his explanations reflects the quality of his understanding.

Gilb, Tom. Principles of Software Engineering Management. Wokingham, England: Addison-Wesley, 1988. Chapters 7 and 15 contain thorough discussions of evolutionary delivery, one of the first incremental development approaches.

Beck, Kent. Extreme Programming Explained: Embrace Change. Reading, MA: Addison-Wesley, 2000. This book contains a more modern, more concise, and more evangelical presentation of many of the ideas in Gilb's book. I personally prefer the depth of analysis presented in Gilb's book, but some readers may find Beck's presentation more accessible or more directly applicable to the kind of project they're working on.