The Greatness.com

« March 2005 | Main | May 2005 »

April 14, 2005

Hilarious Bayesian advertising

Advertising links, or more specifically, the way they are picked to appear on my screen on various Web pages, have really been cracking me up today. CNN's story "Dragnet grabs 10,000 fugitives" served up four text ads, all soliciting for legal services. None of them asked me if I'd been subject to Section 215 of the Patriot Act, but that's probably because they can't help me then. Also, Tech Central Station had another global "variation" article which contained ironic sidebars like "Global warming -- it's happening", from Greenpeace no less. If it were any other publication I'd question their advertising strategy. But since they're already hired guns for Exxon I don't think it would help.

Not related, but troubling: why does anybody care about pregnant Britney?

Posted by The Greatness at 12:37 PM | Comments (0)

April 11, 2005

The Nature of Justification

I promised everybody some thoughts on the nature of justification and how it related to politics. I don't know if anybody cared to see me make good on that promise, but in any event, I'm not sure it matters. Pointless spewing of a blogger's thoughts simply because they are his is, deep down, the raison d'etre of most of the blogosphere. So let's get started!

The best known account of knowledge is "justified true belief" theory, which stipulates that in order to know something (call it P -- all philosophers do),

1. One must have (good?) reason(s) to believe P,
2. One must believe P, and
3. P must be true.

There are some technical problems with this theory as a coherent, exhaustive definition of knowledge that have to do with justification (google 'Gettier' for some examples), but it's close enough for our purposes. We don't have knowledge if we have an unjustified true belief, a justified false belief, justified true unbelief, or some combination of failings like unjustified and false.

For someone to attack a knowledge claim, it is necessary to sever one of these three legs of knowledge. Imagine I assert P: "the sky is red." You know the sky is blue. If you wanted to change my mind, you have a few choices:

• You could tell me, "No, the sky is blue." This strategy -- the direct assault on the truth claim -- doesn't tend to be very successful. Unless I regard you as having superior perception of color or as a great authority on the sky, I'm not likely to abandon my position based simply on your alternative claim.
• You could ask me why I thought the sky was red, in order to ascertain what justification I had for that belief. Chances are I'll say "because that's simply how it appears to me." Further questioning might reveal that I am convinced of the sky's redness because everything looks red. Maybe I have worn glasses coated in red cellophane for my entire life, only taking them off in the dark. In that case it should be possible to change my position simply by making me realize that my glasses deceive me; my reasons for believing the sky is red were not good reasons. But what if I didn't have an obvious sensorial handicap, and I still maintained that I saw a crimson sky? Or what if I took as my reason something I read recently out of the National Midnight Star, a paper I take to be reliable but you find to be of dubious quality? You could talk to me about how something is clearly wrong with my eyesight or my judgment of publication standards, but I still may not budge; indeed, I may have counterarguments for your arguments (I just had my eyes checked, and wasn't TNMS right about Geddy Lee?).
• At this point you are beyond talking to me in a reasonable manner. The sky is not red, it's blue, it always has been and I am not justified in believing otherwise. You begin to think that I'm just deluded, perhaps because believing the sky is blue would cause me to also deny religious truths I hold dear. You pity me for a moment, but then a darker thought enters your mind: I'm just toying with you! I claim that the sky is red and that I have reasons to believe it, but I couldn't really believe it. Why would I do that? You don't know, but all of the reasons you come up with are potentially libelous. Maybe I get a secret glee out of pointless argumentation. Maybe I'm an evangelist for my Red religion who cares less for the truth than for getting new converts. Maybe I'm in the pay of corrupt Redco, who is spewing FD&C Red Number 2 into the atmosphere and wants to get off scot-free. Whatever my motive, it's obviously not truth because THE SKY IS BLUE, DAMMIT!

Note that the rhetorical strategy for convincing someone that something is false if they believe it to be true doesn't differ in any way for any P. I could just as easily have argued against your sapphire worldview using the same methods. The exasperation of someone forced to use the third option is an indication of two things: how hard it can be to change someone's mind when it is made up, and how futile the argumentation can become, full of impotent sound and fury. No matter what I think of such a person, I cannot convince him that P by hurling invective at his belief or his justification.

My example P of a red sky is a bit simplistic on more than one count, of course. Anyone who asserts P can expect to reap a blue whirlwind, and we'd say he deserved it. How could anyone think the sky is anything but blue? Totally unjustified, irrational, nay immoral, that anyone would believe that P! On the other hand, even if there were such a person, we wouldn't waste our time arguing the point; believing such a thing is weird but not necessarily dangerous. Also, a binary notion of belief isn't realistic. I know that we all believe a great many things, each to a different degree and based on varying quantities of Pure Uncompromising Reason^TM, life experiences, intuition, what somebody told me, etc. Some examples on my current continuum of belief:

• That "2 + 2 = 4" for any natural number system of greater than base 3 seems unimpeachable to me.
• I feel serenely confident in the validity of Newtonian gravity as it applies to macroscopic objects traveling much slower than light.
• I believe that Columbus arrived in the Americas in 1492.
• I find evolution to be a very well attested scientific theory to fit the (much firmer) facts of geology and biology.
• I find it entirely plausible that survivors of the Lost Colony intermarried with local tribes.
• I have not decided whether global warming is in large part anthropogenic.
• I am skeptical of superstring theories.
• I'm quite comfortable disbelieving UFO stories.
• And surely the Easter bunny does not exist.

Along this continuum of belief that forms my unique noetic system, there are countless propositions, some consciously placed, others unconsciously sorted. It is in this messy world -- the real one -- that the above rhetorical strategy is usually employed, but to no better effect. Take P to be "speed limits save lives." We can find people who assert P and others who assert ~P in this case. It is an issue that has social value much beyond that of the red/blue sky debate, for persuasion of most people to ~P if P is true will, by extension, cost lives. And most people don't believe P or ~P nearly as strongly as they believe the sky is blue. This is an important issue where reasonable people differ but could be persuaded to change sides if presented with strong evidence for or against. Do we see any consensus emerging from the speed limit question? Far from it! Surf the Net and you'll find "speed kills" zealots who blame the speed-hungry auto industry and the Feds for lack of oversight, as well as unrepentant (and paranoid) speeders who say limits are a front for greedy actuaries and municipalities. Each group has its own favored studies and reasons for why the other studies are wrong, even suspect. Motives of the opposition figure prominently. At times it seems people are more interested in arguing the intrinsic worth (or lack thereof) of speed limits than attempting to chart their consequences.

We hope that when we consider maximally important P's that we will be able to persuade a solid democratic majority and not perish. But isn't that hope as vain in the real world as it was in my red/blue one? Maybe using the standard rhetorical strategy, but not in general, I think. I'm convinced there are effective means of persuasion that do not rely principally on destroying someone's justification or questioning his motives. Methods of persuasion which instead patiently present the evidence and ask: what do you think of P now that you've seen all sides? And count on the other person to be open minded.

No doubt there are things I believe which are false and things which I disbelieve which are true; I hold this meta-belief unswervingly. But which ones? Living a life consistent with truth requires that I regularly expunge false positives and negatives. Alas, a life unbesmirched by falsehood is impossible. I therefore live my life hoping that in dialoging about "big issues" I will fix the major defects on my propositional list, and that in the process some smaller bugs will get squashed. I don't always feel right, but I do feel justified.

Posted by The Greatness at 05:35 PM | Comments (0)

April 04, 2005

malloc madness and Jim Crow

I have serious coding issues today that are really cramping my time. But I felt like making a blog entry anyway, if only to point out something interesting I learned today (which is, of course, the raison d'etre of this blog).

In history class we were taught about Plessy v. Ferguson -- and Jim Crow in general -- as if it were the ineluctable legal consequence of monolithic American racism, where business, government, and majority society were lockstep in support of "separate but equal." We don't tend to doubt this picture because it is held as nearly axiomatic that businesses were run by white men and white men were racist. Change "were" to "are" and you have the substance of every argument against repealing the Civil Rights Act of 1964. But have you ever heard the whole story?

Posted by The Greatness at 01:24 PM | Comments (0)