I can only second this. Especially regarding the 1st Shapiro book: We had an advanced seminar that was essentially "read and discuss the essays in the book". We

I can only second this. Especially regarding the 1st Shapiro book: We had an advanced seminar that was essentially "read and discuss the essays in the book". We were about 6 people, including the (excellent) professor. It was very productive and we got a very good overview of the field back then, and I think everyone enjoyed that seminar very much. To relax from the reading, and since most (all?) of us did already have some advanced math education beyond the average philosophy student (e.g. I did major in CS, others had maths major or minor), we did a "guided by the professor" proof for the incompleteness theorem - he always wanted to do that, but couldn't in his normal lectures. That was fun. I'd suspect I wouldn't have had such fond memories if we everyone was to read the original sources, especially since that would have been nearly impossible in the given time frame. In most seminars that's usually accelerated by having one person read one source each and do a presentation and/or write an essay. I think reading a well written essay and discussing that in a small group is the superior approach. Of course if you want to become a professional in the field this will not replace reading some/most/all of the original sources in the long term. It is then a matter of opinion if it's better to start reading the source with or without some broad-but-shallow knowledge of the field. I prefer with, since I never read such material to "just learn what X said on Y so I pass some exam", but because I like thinking about it and put it into perspective [e.g. temporal, but also alternate ways of looking at things]; but then I'm not a professional in that field, so insert shrug-pony here.