Rule 3: Suspect the text
As I've frequently emphasized on the podcast, texts often have a long and complicated history of transmission. A work by, say, Aristotle was first written down well over two millenia ago; it's not unlikely that even the very first copy/copies had mistakes, given that it would presumably have been dictated to a scribe. To reach us, it then had to be copied by hand many many times, with the earliest surviving copies being copies of copies of copies... And those earliest surviving copies come from the Byzantine period, many centuries after Aristotle. Of course things aren't quite so daunting with more recent works but certainly anything produced before the invention of printing involves this kind of transmission of copying, and there are philological issues to contend with even in the case of early printed works. This means that, if you are really getting into the nitty gritty of a pre-modern philosophical text, you need to beware of the existence of many variants in the text, which could radically alter the meaning. Scribes made mistakes, incorporated glosses into the main text, and made their own emendations to fix problems they found in their copies (these scribes were not stupid by the way: their emendations may well be right!). And that isn't even taking into account the possibility of outright tampering. The podcast fell afoul of this recently when I emphasized the salacious story about Avicenna's unrestrained sexual appetite while dying of colic; then I became aware of a recent article showing persuasively that this was a later, hostile addition to the biography of Avicenna written by one of his students. (See the comments on that episode, here.) The upshot is that historians of philosophy need to be philologists too, insofar as they can manage it, and to take seriously the work of scholars working on textual transmission or even collaborate with them.