This is all well and good for certain types of notes, such as summaries of books that I read for exams. It’s kind of like scribbling marginalia in a book. But this method is not so good for other types of notes, such as pieces of evidence or quotations. The problem is that attaching notes to the source forces you to think about the source first and then the idea encapsulated in the note, rather than the other way round.

I first learned how to take scholarly notes on index cards: one thought per card, with carefully marked keys to subjects and sources. I still think that index cards have some virtues that digital note-taking can’t beat. What I wanted from Zotero was a way to think about notes that was more like the model of index cards and less like the model of marginalia.

Lo and behold, Zotero had the necessary functionality for years. The problem was not the capabilities of the software, but the way I was thinking about taking notes.

What I do now is make a standalone note for each thought or piece of evidence. But I also make the note a related item of the source from which I got the idea. You can see in the screenshot below that this standalone note is related to a book.