Pages Menu
TwitterRssFacebook
Categories Menu

Posted by on Feb 28, 2014 in Uncategorized | 4 comments

Right Theory, Wrong Theory

Almost any event that is happening now or has happened in the past can be explained more than one way. We could explain the assassination of John F. Kennedy with the theory that he was shot by Lee Harvey Oswald alone, by his wife Jaqueline Kennedy, by the CIA, and so on and so forth.

How do we tell which theory is right? There are three main factors that add (or subtract) likelihood:

Ad hocness. How many assumptions the theory makes. A theory that makes a lot of assumptions is less likely than a theory that only makes a few assumptions. Of course, there are exceptions: If we’ve got one theory that makes several believable assumptions and a theory that makes a really unbelievable assumption, we’d go the believable set of assumptions. If we have a theory that makes more assumptions but can explain a lot more of the facts, we might go with the theory that explains more. It all depends.

Inductive support. A theory that is supported by generalized statements that have proven true in dozens/hundreds/thousands/millions of cases is better off than a theory that is not. A theory contradicted by generalized statements that have proven true in dozens/hundreds/thousands/millions of cases is at an extreme disadvantage. Example: there’s a story about Merlin the wizard moving the stones in their places at stone henge with his magic. Since we no evidence of magic in other cases (meaning that the existence of magic is not supported by any true generalizations) and lots of evidence that people make up stories (“stories about magic are usually just legends” is a true generalization about the way the world works) that means that Merlin-magic theory lacks inductive support, while the belief that the story is a legend has lots of inductive support, and that’s a good reason to believe one over the other.

Fit with the Evidence. This one’s easy: just look at the facts you’ve got and the probability each theory gives to those facts. Example: if the theory of evolution is true, it is one hundred percent likely that intermediates between ancient apes and modern human beings once existed. If the theory of evolution is false, it is less than one hundred percent likely that they did, simply because there’s no reason to believe such animals ever existed if evolution is not true. Comparing how well theories fit a large range of facts can be pretty complicated. Sometimes it’s really easy and straight-forward, but… not all the time. There is a way to unwrangle the complexity of evidence-fitting when it’s needed, but I won’t go into it here (hint: Bayes’ Theorem).

Not one of these criteria is absolute. It’s like a point system. Explaining one fact may not win the game for a theory, but it will add a point or two to the scorecard.

As a test case, let’s look at Kris Komarnitsky’s hypothesis about early Christianity (disclosure: understanding this may require reading the book). Komarnitsky proposes something along these lines:

“After Jesus died, he was given a common ground burial in a place unknown to his followers. In the wake of Jesus’ death, his followers had a great need to make sense of or rationalize what had happened. This need, combined with a few hallucinations of the risen Jesus that some of them had. As these beliefs were passed on, the process of legendary embellishment worked its magic on Christian beliefs, and from this came the gospel story that Jesus’ tomb had been discovered empty.”

How many postulates are here? I’ve underlined everything that might be a discrete postulate and added a number to it:

“After Jesus died, he (1) was given a common ground burial in a place (2) unknown to his followers. In the wake of Jesus’ death, (3) his followers had a great need to make sense of or rationalize what had happened. This need, combined with (4) a few hallucinations of the risen Jesus that some of them had, (5) spawned a belief that Jesus had been raised into heaven. As these beliefs were passed on, (6) the process of legendary embellishment worked its magic on Christian beliefs, and from this came the gospel story that Jesus’ tomb had been discovered empty.

Komarnitsky demonstrates that (1) (“Jesus was given a common ground burial”) has modestly strong inductive support: there’s plenty of written and archaeological evidence showing that most criminals were buried in the ground, not in a tomb. It has some explanatory power: it explains why some early Christians thought Jesus was buried in the sand (see “Secret Gospel of James) though it does not explain why there is a story about Jesus being buried in a tomb, which is why Komarnitsky needs his sixth postulate.

(2) has some reasonable inductive support and a small amount of evidential support (see chapter 2, Doubting Jesus’ Resurrection).

(3) has inductive support, since we can expect this would have happened based on what happens with other cults.

(4) has inductive support, since we know people commonly hallucinate after the death of a loved one. Richard Carrier wonders why Jesus did not appear to everyone all over the world, or to any enemies of Christianity besides “one odd fellow” (Paul) and sees this as a strong evidential consideration in favor of the hallucination hypothesis and against the resurrection.

(5) is not so much a postulate as an assertion that Komarnitsky’s hypothesis could produce all the facts. I don’t know of any inductive or evidential considerations that bear on this (although, again, this isn’t a postulate and so it shouldn’t be evaluated as such). However, given that (1) people would have had hallucinations of Jesus after his death, and (2) the fact that early Christians may have thought the resurrection was near / were familiar with the concept of a dying and rising son of god, it is extremely plausible to believe somebody would have put two and two together and created the resurrection belief.

(6) is inductively supported in multiple ways. First, we know myth-making happens generally among all groups of humans. Second, we can prove legendary embellishment happened within the Christian community specifically. Take a look at the gospel of Thomas or the gospel of Peter and just look at the legends that cropped up. Or compare the gospel of John (thought to be written last by most scholars, including evagelical Christians) with what the apostle Paul wrote, or what the synoptic gospels say. We also have good evidential support for this: Paul did not mention the empty tomb when he argued for the resurrection of Christ in one of the earliest documents we have about Christianity (1 Corinthians 15). This is damningly difficult for the resurrection theory: almost every Christian every where has mentioned the empty tomb when arguing for Christ’s resurrection. Why didn’t Paul? Answer: the empty tomb story is a myth fabricated by later Christians.

In short, Komarnitsky’s overall theory (with its five postulates) explains four facts that I find significant and highly supportive, plus a few other facts that offer minor support. It explains the empty tomb story and the appearance traditions (that’s two), and it also explains (1) why Paul never mentioned an empty tomb and (2) why the resurrected Jesus was reportedly seen in just one culture and time and not all over the world. That’s four facts.

Let’s compare Komarnitsky’s theory to the standard Christian theory:

(1) God exists. (2) God can and will intervene in the workings of nature. (3) God intervened in nature on the specific occasion of Jesus’ death to raise him from the dead according to some desire God has for the world.

(1) and (2) obviously affect the likelihood of the whole theory, and to the extent that one doubts either of those he must also doubt the third assumption at least as much. However, I’m going to set those two assumptions aside and critique the Christian theory purely on the basis of (3).

(3) doesn’t have any positive inductive support: you cannot make well-tested generalizations about how God intervenes in nature or about God’s desires. The desires of God may be open to philosophical speculation or the personal “will to believe” but the desires of God (if there is a God) cannot be described as “fact” or even as a good theory explaining patterns in the facts we have. In fact, there is a really good inductive argument against (3): we’ve got tons of evidence to prove that most miracle claims are false (I’ve argued this extensively here and here).

Proponents of (3) argue that it has some explanatory merit: it can explain why the gospels tell a story about Jesus’ tomb being empty (because it really happened and tomb’s emptiness was a side-effect of Jesus being resurrected and leaving the tomb) and that some early Christians said they saw Jesus after his death. I agree: it can explain these facts. But is it the best explanation? Since I’ve agreed to set aside postulates (1) and (2), let’s compare our two theories side by side:

Postulates Inductive Support? How Many Facts Explained?
Komarnitsky’s Theory Five Modestly Supported           Four
Christian Theory One* Strong Inductive argument Against            Two

 

Here’s where you get a voice: What do you think about this table? Accurate or inaccurate? If you agree with the table, what conclusion do you draw from it?

*Actually three, but I said I’d be setting aside the question of God’s existence and will to perform miracles, so I only counted one.

  • SmilodonsRetreat

    I would submit that in almost every case, the word “theory” should be replaced with “hypothesis” in post.

    You’ve got one place (A theory that is supported by generalized statements that have proven true in dozens/hundreds/thousands/millions of cases is better off than a theory that is not. – See more at: http://www.skepticink.com/humesapprentice/2014/02/28/right-theory-wrong-theory/#sthash.GL39si60.dpuf) where theory is used in two different ways.

    Arguably, you’re using the casual sense of the word, not the scientific sense. But I think it would still be better to use hypothesis in most of the instances.

    • ncovington89

      The word “theory” is used to mean the same thing in the sentence you quoted.

      Let me explain the meaning of that sentence: A theory/hypothesis/explanation, like, say, the theory that I was formed by sexual reproduction, is more credible than another theory (like, say, that I was formed by spontaneous generation) because the sex theory has inductive support: we know from innumerable experiences that living beings are created through sex, zero observed cases where they aren’t, and that supports the sex theory.

      • SmilodonsRetreat

        Again, this me just being a nitpicker about this (since the recent discussion at Debunking Christianity). If it doesn’t have a lot of evidential support, then it’s not a theory. It’s a hypothesis. In this case, sex is a theory, because, as you say, we have a lot of evidence to support it. Spontaneous generation isn’t a theory, it’s a failed hypothesis.

        I’m just concerned because you do talk about both hypotheses and theories in the same article using the same terms. That’s what I’m trying to get across.

        • ncovington89

          You’re right. I throw the word “theory” around pretty loosely. My definition of “theory” is synonymous with “explanation.” Scientists and historians don’t always use the word consistently, though I do understand that some use the word in a different sense than a mere “explanation” and I also understand that there is some value in doing that. I’m a little sloppy with my English :-)