Notes
The craft secret of scholarship is to take notes on what you read. Keep in mind that you are preparing these notes as a gift to yourself in the future. In this future situation you will no longer remember anything about the book, other than that you read it. You should consider layout--for easy scanning. Don't pack the sentences/phrases into dense clumps, thinking that the first priority is to save space. Page numbers in the left margin let you know at once if you have skipped pages, and give an overview of where the information is relative to the whole. Using the vertical spread, separate idea on separate lines, allows you to scan the notes to locate important organizing ideas. You know well how argumentation works, and the relation of outlining to "proof." The notes do not reduce neatly to an annotated outline, but that idea of a linearly unfolding demonstration is at work in any book. Notes enforce/enable a level of attention that is impossible to sustain otherwise. No need to try to understand the book as you go along (an advantage with theoretical texts). The overall sense emerges in retrospect, when you review your notes.
The purpose of the notes is to capture a good level of detail, but not to interpret. The notes are an annotated index, faithful to the book rather than to any current project. Comments are ok as long as they are inscribed in a different colored ink (or distinguished by font if you are typing). Some people use shorthand and then type the results, but that takes immense discipline. Those who are used to the computer can type the notes directly as they read, and then use spreadsheet methods to find patterns. If you are handwriting the notes, you might want to consider making that process the one set of notes you will make, and do it so they are legible as is, without typing. Keep in mind that if you are serious about this method you will be doing hundreds of these sets of notes. of course you don't do notes on every book, but only the difficult theory books or scholarly studies central to a project for which you need control of the information.