Thursday, February 16, 2006

This Better Be Good

In Aristotle's Spiral, Ernie has begun to describe his epistemology, through which he claims to be able to know that Christianity is true, for appropriate definitions of "know", "Christianity" and "true". (A clever statement I heard once says "One is nearly equal to two, for large values of one and small values of two.") He hopes that we will be able to "define a mutually-satisfying epistemic framework" which has started with our agreement on two axiomatic principles:

  1. Truth exists
  2. Belief in Truth is Good
In my last three posts, I tried to explain the difficulty I had seeing how Ernie arrived at belief in Christianity from a small number of generic definitions and axioms, and particularly how he could defend his answer to my question about what could lead him (roughly speaking) to abandon belief in Christianity. As I suspected, Ernie has confirmed that there are additional supporting "inputs" and "identifying those is the goal of [Ernie's] epistemic framework." As we identify them, we ought also be able to identify further possibilities for falsification. (Do you agree, Ernie?)

After some lame excuse about not being able to solve ancient philosophical puzzles in a single post :-), he offers the following additional epistemological principles:

  1. Knowledge imperfectly reflects Truth
  2. Better Knowledge produces Better results
  3. Knowledges improves through honest collaborative inquiry into shared Reality

Do I agree with these? Well, I have no problem with (3). (4) needs a bit of clarification. "Better knowledge" is, I presume, knowledge that more perfectly reflects truth. But what are "better results"? And can knowledge itself produce anything?

My explanation of "better results" would be something like this: "Better results are results that more closely match the intended results of an agent's action." There is a more general principle that might be stated that emphasizes the predictive aspect and de-emphasizes the agent's involvement: "Better knowledge produces better predictions." (In either principle, "produces" should be understood rather loosely, as the knowledge does not itself produce anything; instead it is used to produce something.)

I am uncertain, however, if this is what Ernie intended by "better results". I will return to this later where it becomes more clear.

Regarding (5), I agree with the statement itself, but I do not agree with some of Ernie's explanation:

That is, I reject both the Platonic idea that knowledge is truth, and the postmodern idea that knowledge is disconnected from truth. Rather, I explicitly affirm both Community and Reality as the critical inputs necessary to enable my Knowledge to asymptotically approach Truth. In spirit, I often find myself much closer to Aristotle than any other school.

While community is very helpful in developing knowledge, I am not so sure that it is critical, nor am I convinced that (as Ernie later says) "Knowledge, to be validated, must be encoded in a form accessible by a given community." Still, at this point I am unsure how important this principle is to our discussion, so I am not going to get hung up on this. As far as reality being a critical input, I wonder if I understand what Ernie is driving at there, as this seems true almost by definition. Loosely speaking, I would call "truth" the perfect description of reality. In this sense, reality is obviously necessary for truth to exist, and is therefore a requirement for knowledge as well. But that does not seem to help too much. Am I missing an essential distinction?

Ernie lists four implications of his epistemology:

  1. The purpose of these axiomatic truths is to enable greater knowledge
  2. Knowledge, to be validated, must be encoded in a form accessible by a given community
  3. Knowledge must therefore be tested against a reasonable shared reality
  4. The ultimate test of the quality of knowledge is whether it produces Good outcomes

As before, we might quibble over some details but I think the critical issues that need to be addressed are raised by (d). This is part of what leads me to wonder about Ernie's earlier use of "better results". Here, "good outcomes" seems to be a description of moral worth, and this becomes even more clear later when Ernie says "The Truth is what Works" (but "What Works is not the Truth") and later explains

Note that "Works" in this context means "successfully creates Good." Thus, it is important to note that my epistemology is incomplete without a companion ethical framework, which is logically the next thing on the agenda.

Based on these later statements, I believe Ernie is trying to tie knowledge to moral good, such that you can uncover truth by looking at what produces moral good. This presents several difficulties. First, what you consider to be morally good may very well depend on what you believe is true, so that you risk being trapped by your prior beliefs. Second, measuring the total moral good (or evil) associated with some belief or action strikes me as difficult; causes and effects are hard to untangle and the potential for bias is high. These are serious problems.

But even before you get to those two problems that would be encountered if we accepted this causal relationship between knowledge and moral good, there is a still more important difficulty. Is this principle reasonable? Is the causal relationship valid? I think not. Like most tools, knowledge can be used for good or for evil. We might try to rescue the principle by saying "The ultimate test of the quality of knowledge is whether moral agents are able to effect good outcomes." But even so, this seems a rather gratuitous addition to the epistemology. There are many kinds of knowledge that are entirely amoral in nature. To state that the ultimate test of knowledge is based on good moral outcomes is reaching too far.

As far as I can tell, this principle appears to be central to Ernie's epistemology. For instance, consider these additional statements he makes:

  1. My goal is to do as much Good as possible.
  2. In order to do (1), I must believe as much Truth as possible
  3. This requires discovering what beliefs Work to create Good
  4. Those beliefs must simultaneously motivate personal, social, and intellectual virtue

These statements reflect, I think, the importance Ernie places on the "knowledge leads to good" (KLTG) principle.

There are a few more things that I should touch on, but this is enough for tonight, except for this bit of humor. Search for "epistemology" on Google. The only ad that comes up (for me at least) is from e-Bay: "Looking for Epistemology? Find exactly what you want today." Why didn't I think of that before?

No comments: