101 Interesting Things, part fifty-four (a): The WizKID Epistemic Framework - Intro & Advantages

Philosophers as a discipline are still stuck on the Justified True Belief (JTB) model of knowledge, even though Gettier showed pretty effectively where that breaks down.  But even better, information scientists have made a replacement model that works, is quantifiable, and more accurately reflects how we construct and use knowledge in real life.  I hope that one day philosophers manage to bring this better model over to their side, but I'm not interested in doing that because I'd have to fight against the tide and I'd really rather just do my own productive work.

But if you wanna know how knowledge really works, and get in on the ground floor of cutting-edge epistemology, then have I got something for you!

Information scientists have a framework that starts with data, contextualizes that into information, relates that into knowledge, and distills the whole thing into wisdom.  It's called various things, but the acronyms tend to be unfortunate, so I like to go top-down and call it the WizKID framework (I also don't like calling it a pyramid, for reasons I'll discuss later).  Here are the basics of the relationships between the different levels:

• Data is "just numbers," like 1 or 3.
• Information is data with context, like 1 gallon or 3 dollars.
• Knowledge is information that's been interrelated to other information; for example, "1 gallon of milk costs 3 dollars American."
• Wisdom is a distillation of various knowledges into something actionable, like whether $3/gallon is a good price for milk.

Now, I bet if you think real hard, you could come up with all that follows.  I mean, you're smart.  (You read my blog, after all.)  But I've been thinking pretty hard about it for years, so I wanted to share a few of the most interesting things that come out of the above, organized by the following loose headings:  advantages over the JTB model (for today), and theoretical & practical consequences (for later).

Advantages of WizKID over JTB:
Why I Left the Philosophy Department

Gettier problems aren't the only problems with the JTB; for one thing, it absolutely reeks of logical positivism.  But every component of the JTB is fraught all on its own!

Justification concerns what "should" count as a reason for believing something - "I proved it with math," "I saw it with my own eyes," "Teacher said so," "We got five sigma verification in subsequent trials," these are all common justifications for believing something is so.  What's more, they can all be good ones, even if they're not foolproof or context-independent:  math is great for deductive truths because mathematics is the study of interesting tautologies; our senses are unreliable with disconcerting frequency, but pretty reliable for most situations and often all we have to go on; teachers are the authority on classroom rules, and their whole job is to educate kids, so they're probably a pretty safe bet (until you're older, anyway); and demanding statistical confidence intervals, while hard to come by, are often what you need when you're dealing with weird shit.  But sometimes these justifications are flimsy, inappropriate, or outdated, so the fact that we've got a justification doesn't really tell us much unless it holds up.

Truth is... just... weird.  It's one of those things where everyone thinks it's obvious (or at least simple), but really there's a whole lot to unpack.  But at the end of the day, "truth" is a word, and words are made-up.  Moreover, it's a word we use to talk about other words, so the made-up-ness is even stronger (and gets even sketchier when we get into problems related to Russell's paradox, where true and false look non-binary because you can get values that look like "both" and "neither").  And finally, truth gives us second-order knowledge problems ("how do you know that you know..."), because if something needs to be true for us to know it, then we can't establish its truth without bringing in another truth assessment - in other words, to know if something is true requires having a justified true belief about the first belief's truth.  You can say, "Well, true things are true, whether we know it or not," but that just puts knowledge out of useful reach, so who cares?  Ugh, I hate talking about truth.

Belief is probably the most innocuous one, because it stacks up pretty well with most non-philosophers' common sense to say, "You believe the things you know."  Except that we don't always:  we often find "implicit knowledge," things we act as if we know without really believing them.  These can be hard to come up with, because once you point out an unexamined bit of knowledge, then poof it's been examined and the person decides to believe it or not.  But moreover, the need for a belief to be held (regardless of whether justification and truth can be effectively assessed) makes knowledge propositional in nature.  So philosophers then had to invent the distinction between propositional knowledge (which can be put into sentences) and procedural knowledge (which can't, such as knowledge of how to ride a bike).  Procedural knowledge can be loosely translated into propositions, for sure - that's the whole conceit of instructions, after all - but someone can know how to ride a bike without ever putting that knowledge into believable propositions which have justification and truth value.

WizKID doesn't have any of these problems, because it doesn't construct knowledge in this way.  What's more, it more accurately reflects how we actually use knowledge:

I mean, it's much more vivid in my head, but the randomness is definitely there.

Because relationships between information constitute knowledge (rather than entire statements constituting knowledge), this makes knowledge "webs" easier to manipulate or modify:  we change the relationship, or swap out a thing for another thing.  Sometimes we get... ::shudder:: "floating abstractions," as Ayn "Even a broken clock is right twice a day" Rand called them:  something that's not firmly attached to anything, just kind of "out there," which is how all knowledge webs start and how you can compartmentalize different kinds of knowledge from each other.

Now, I want to be clear here that WizKID has problems:  it's a damn sight better at integrating procedural knowledge into the overall framework than JTB, but it still goes squiffy in the particulars; it doesn't get us to "true knowledge" as JTB proponents purport (or at least want) to do; and it's considerably messier, because a knowledge claim can't be analyzed in isolation but only in context of an existing knowledge web.  There are others, to be sure, but notably:  all the weaknesses of WizKID are also places where it's still better than JTB.

That's enough for now, because the theoretical section is pretty damn long and we got pretty heavy already, so I'm splitting this entry into two posts.  Ta!