Good points about scientific writing

In Richard Preston's introduction for The Best American Science and Nature Writing 2007, I found some good points about scientific writing.

• Writing is linear, proceeding from one word to the next; one of the tricks in science and nature writing, which works well in certain kind of pieces but is a hard trick to get right, is the delicate process of managing the line of prose as it moves from exposition to narrative back to exposition again. Technically, when you break a narrative to explain something, it's called a "set piece." The name tells what it does. It sits there, providing reader with explanation of something. If you put too much narrative in a piece of writing about science, without enough exposition, the reader won't see the reason for the narrative. But if you start a narrative and then hang too much exposition on it, the exposition ends up as a load of wet laundry hanging on the line, and it drags the narrative down to the ground.
• Another thing I relish is clear exposition of an important idea, especially if it's counterintuitive, challenging, controversial, or hasn't been presented in such a way before. Here we don't need narrative or character, what we need is a good argument. "A Plan to Keep Carbon in Check," in Scientific American, by Robert H. Socolow and Stephen W. Pacala, who are both working scientists, has a quiet but strong voice and a good raison d'etre. The authors ask the straightforward question, How, practically speaking, can the world reduce carbon emissions? If we know the climate is warming up because of human-caused carbon emissions, then if we begin acting right now, what can we do to reduce carbon emissions, and how effective will is be? In a straightforward, persuasive set of arguments, Socolow and Pacala show that carbon emissions can be lowered, and it can have a major effect.