...Stuff publicly available on on arxiv, pubmed etc. I quite often dig around in google scholar but i really don't stand a chance of interpreting how relevant any particular paper is to it's field. 'Summary' papers can be quite enlightening in that they outline where stuff might potentially be headed, but even then... maybe i'd probably be better off sticking to textbooks (or even wikipedia). Also, when does the focus of science education shift from studying textbooks to studying the papers the textbooks are based on? Or do you go straight from textbooks to new research papers? I suppose there must be textbooks written specifically for phd students...just wondering how much people read papers before they specialize and outside of what they specialize in.