Science and rationalizationMarch 26, 2009 — Deacon Duncan
In a comment on yesterday’s post, Jayman raises a very good question.
DD, I don’t see why additional information about ghosts is necessary to test my hypothesis. If we identified a ghost as a deceased person my hypothesis would be confirmed. It doesn’t matter whether you would still have additional questions about ghosts or souls or spirits.
Ok, so it’s not exactly phrased as a question, but the implication is there. Why isn’t the test, taken in isolation and without regard to other factors, sufficient to establish the hypothesis? It’s a good question and it points up an important principle that I neglected to cover in yesterday’s post.
Let’s say that I’m a little short on cash, and so I decide to go into the diet pill business. I make up a large batch of pills, and offer free samples to anyone who wants to lose weight. My pills, I tell people, don’t magically make fat disappear. Instead, they work with your body to multiply the effects of moderate exercise and a bit of portion control in your eating. I get about 100 people to try my free samples, and I predict that if my pills really work, I ought to see people losing weight when they take it.
So I check up on my free sample recipients and find that quite a lot of them experience some weight loss while taking my pill. In most cases it’s more a fluctuation than a steady decline, but at least part of the fluctuation is loss, right? And a smaller number actually do experience the steady decline, with a handful of soon-to-appear-in-my-advertisements people who experience dramatic weight loss. So my tests prove that my pills really work, right? Did I mention that they were just cornstarch and water, formed into tablets and baked until hard?
My experiment fails because I failed to control for other variables. I had a biased sample to begin with: people who wanted to lose weight. I created a context in which they were likely to employ other weight-loss methods (diet and exercise). And my test was designed to single out successful results while discounting the failures (“individual results vary”). Whether by intent or by accident, I created a “test” that produced the biased conclusion I wanted to reach, rather than arriving at the unbiased truth.
So the answer to Jayman’s implicit question is that we want to look at the situation from all angles, and make sure that we’re not just creating a pseudo-scientific excuse for jumping to a predetermined conclusion. Our goal, as skeptics and scientists, is not to try and find some way to confirm someone’s opinion, but rather to discover what the truth really is. That means we want to apply rigorous tests and not just informal assays.
Turn it around just a bit: if our goal is to have a solid, reasonable basis for our conclusions, why would we want to rule out the additional questions about spirits (to return to Jayman’s original hypothesis)? Why would we want to forbid certain questions from being asked? Why would we want to insist on drawing our conclusions before we find out whether the “spirits = ghosts” hypothesis is really consistent with itself and with the real world evidence? If it turns out that spirits don’t actually exist, shouldn’t that have a significant impact on how we interpret the results of our test?
Remember, our core principle is that truth is consistent with itself, and one of the implications of that principle is that when our beliefs are untrue, they’re going to conflict with the real world evidence. If we invent rationalizations to try and explain away the inconsistencies, we may succeed in creating an apparent reconciliation in one specific area, but since the rationalization is untrue, it’s going to create new inconsistencies in other areas. Thus, to know whether we are uncovering new truth, or merely covering up an untruth with a plausible (but untrue) rationalization, we need to explore these other implications of our premises.
My goal, as a Christian, was simple. I knew that different men said different things about God. All I wanted was a reliable means of determining which of those men, if any, were really telling the truth about Him. I didn’t want to simply put my faith in whatever men said was right, even if (especially if!) the man I was trusting was myself. I knew the folly of believing whatever seems right in one’s own eyes, and I eventually learned that this was no less an unreliable source if you transformed it slightly by turning it into “whatever interpretation of Scripture seems right in one’s own eyes.” I wanted to know the real truth, the truth that was not built out of the things men thought were right and wanted to believe. The truth that exists on its own, independently of the beliefs of men.
And yet, despite my good intentions, I deceived myself for decades. I made exceptions. I assumed that the men who wrote the Bible were necessarily telling the truth, and that the men who canonized the Scripture were necessarily correct. After all, God would not allow a false book to bear the name of “God’s Word,” right? I told myself I was being objective and verifying my beliefs when what I was actually doing was setting up isolated little self-contained assays designed to reinforce this or that preconceived idea I was having doubts about. I kept my attention focused on the small picture, so that I would not be troubled by the inconsistencies that arose when you try and put all the little pieces together in one big picture.
And it’s the big picture that gives Christianity problems. I once helped a Mormon lady deconvert from Mormonism by the same approach. Individually, the little pieces of her faith were not a problem, and she had a million and one little tests by which she knew that the LDS church was the One True Faith. When I exposed her to the big picture, though, she started to have some doubts. I showed her some of the contradictions in her faith, in front of her own Mormon elders (a couple 18-year-olds), and she began to realize that it didn’t really all fit together. And, ironically enough, my own faith suffered a similar fate starting a few years later.
If our goal is to merely reinforce our preconceived ideas and to insulate ourselves from real-world truth, then fine, it’s ok to wall off those other, potentially troublesome questions, and just limit ourselves to simple assays that will easily satisfy our desire to claim to have some real-world support for our dogmas. But if our goal is to challenge ourselves, and make reality the standard by which we measure our beliefs instead of vice versa, then we’ll be eager to explore all the questions, and to see whether or not our conclusions really are consistent with the real world, even in areas outside our “little tests.”
Truth is consistent with itself, both in the fact that it does not contradict itself and in the fact that each real world truth is interrelated with other real world truths, such that we can follow the connections to discover new truths on the basis of old ones. If we’re not exploiting these very useful properties of the truth, if we’re reluctant to even try to follow all of the ramifications, maybe we’re trying to tell ourselves something. Maybe we’re not really as fond of the truth as we’d like to think. But that’s a human frailty, and the cure is simple: embrace the truth anyway. In the long run, that’s by far the best approach.