Back to Basics in the Law of Evidence - but How?

Systematic treatises on the law of evidence -- the few that exist -- generally begin with (an attempt at) a definition of "evidence." The definitions I have seen -- I have often complained -- are unsatisfactory. Well, OK, so they are. But how does one do better? Perhaps one way to make a start at doing better is to be more clear about the definition of "definition": What makes a definition good and what makes a definition defective? A post by Lawrence Solum (April 11, 2008) on his blog led me to the Stanford Encyclopedia of Philosophy: Prof. Solum had noted that Anil Gupta had posted new entry (April 10, 2008) in the online Stanford Encyclopedia of Philosophy on "definitions."

Reading Gupta's entry, however, was a bracing experience. Although I have no degree in philosophy, I have been interested in philosophy, logic, epistemology, and similar matters for quite some time. Even so, Gupta's entry was plainly not written for the likes of me. I could -- with considerable labor --, I could make some sense -- a little bit of sense -- out of what he said; i.e., I could get the general drift of some of his arguments and analyses. (As far as I could tell, Gupta was in fact saying sensible and intelligent things.) But I could not follow the details of the arguments.

I have some modest familiarity with symbolic logic. But the notation Gupta uses goes beyond my ken. Perhaps this is because I never knew my symbolic logic well enough to begin with. Perhaps it is partly because symbolic notation keeps changing and an autodidact cannot hope to keep up with such changes. Whatever. I am left with a problem, a question: Where (dear Reader) can I find an intelligent, modern, philosophically-sophisticated, and "accessible" discussion of the nature of definition? Or, dear Reader, would you advise me to avoid the problem of defining defining and talk instead (simply?) about the "concept" (rather than "definition") of evidence? If I did that (in a treatise, let us say), would I really have accomplished (or avoided) anything? I humbly await your opinion, dear Reader.

N.B. If one is in a philosophical mood -- and I often am -- one good place to start might be Peter Achinstein's The Book of Evidence (2001). But Achinstein pays little attention to law or to the role or nature of evidence outside of the hard sciences. So perhaps it would be more advisable for me to get my bearings by reading and studying further James Franklin's The Science of Conjecture (2002).