Hi Brian,

I don't think the definition is good (of course, I guess that all depends on how 'justified' is defined!) since it carries deontological, internalist, and/or evidentialist baggage.

Thus, following Plantinga, I'd opt for 'warrant' as that elusive property that turns a mere true belief into knowledge. (Unless you're using the terms interchangeably.)

Obviously just holding a true belief isn't sufficient for knowledge since we wouldn't be able to rule out coincidence or lucky guesses. And who wants to call a lucky guess, 'knowledge?'

I think Plantinga has demonstrated the deontological, internalist, and evidentialist constraints tied up in the notion of 'justification' (see esp. Warrant: The Current Debate, and Warrant and Proper Function).

So, as in much of the literature, a 'justified belief' is a belief you are entitled to hold. One flouts no epistemic duties in holding it. You are within your 'epistemic rights' in holding it. Or, given the rest of what we believe, they fit within our ken. That is, they submit to the standards of what we may reasonably believe. To this Audi adds, "That, in turn, is roughly what we may believe without being subject to certain kinds of criticisms, say as intellectually lax, sloppy, overhasty, or the like" (Epistemology: A Contemporary Introduction to the Theory of Knowledge, p.2). Note the deontologism.

But since this term has been debated vigorously, I don't see any conclusion being reached on the PB.

And, there are some who add 4th conditions. So, justified, true belief isn't enough. The belief must be formed by a reliable cognitive faculty. But this doesn't rule out cognitive malfunction, which seems to be a necessary component for knowledge (ruling out malfunction, not malfunction, that is!).

Furthermore, not only do epistemologists debate what 'justification' means, some say you must also have the right kind of justification, e.g., a belief that is indefeasible (cf. Leher). There are positive and negative defeater conditions too.

Anyway, to answer the basic question, it appears that a survey of the literature tells us that a 'justified belief' is one that has deontological, internalist, and/or evidential requirement associated with it. Since I'm an externalist (with a negative account of defeat, thus intimating an internalist flair), and a proper function epistemologist (with the caveats of a good design plan, a design plan aimed at producing mostly true beliefs, and being in the right environment, i.e., one sufficiently similar to those my cognitive faculties were made to function in), I obviously take issue with how 'justification' is normally understood. I admit people can be justified in many of their beliefs (even someone deceived by an evil demon who nevertheless does his epistemic duty), yet that doesn't mean that they have knowledge (even if the belief is justified and true).

In answer to Civbert's def, I take issue thusly:

Justification = df 'justified' is deducible from prior true premises.

Seems to me this has numerous problems, not the least of which is an invitation to an infinite regress.

If 'justified' = 'deducible from prior true premises' then we must ask if this belief is justified? Either it is, or it isn't. If it isn't, why should we take his definition? If it isn't, then where are the 'true premises' he deduced this definition from. And, say he gives those premises. Are they also justified? If not, then why take them as offering support for an answer to our first question? Do they have any positive epistemic status, at all? If so, then what 'true premises' are these ones deduced from. And on and on...

__________________
Regards,

J.J.
PCA
Suffix