We've already seen that having only common sense guide the code inspection leads to problems in the checking phase. Because the differing ideas of what makes sense, what is important and what can cause real problems meet in the feedback phase, it is in this phase that the real friction occurs. The people don’t even have to meet, it is enough that they have to transfer conflicting information. In this kind of informational clash, the checkers might feel strongly about certain findings, which the author doesn't consider important enough to note.

In the feedback phase, the differences between people either amplify or are ironed out. Neither action is necessarily feasible to the inspection. A code inspection looking for consensus isn't an optimal one as discussions are often a waste of time. On the other hand, letting people fight for their findings or defending their product isn't good for the team spirit. Implicit guidelines are not optimal, no matter how common sense it all is. You cannot direct cooperation with private, implicit guidelines. They provide no guidance, the whole process is based on the assumption that people share enough of their common sense to have common ground.

Fighting for findings or touchy-feely general opinion seeking isn't practical. Common sense guidelines make code inspections sensitive, not sensible!

Chapter 2. "Symptoms" describes the psychological setup of inspection. Common sense - as will be clear in other parts of the book, too - makes no sense in a collaborative undertaking aimed at finding the most defects possible. In other words, if you can define clear rules for steering some behaviour, why rely on word-of-mouth?