Friday, January 22, 2016

Confirmation bias

Confirmation bias is one of the major logical fallacies. When philosopher Chris DiCarlo and I were teaching a course on critical thinking we used to spend quite a bit of time on it because it's a very common trap. We are all guilty, from time to time, of focusing on just the evidence that confirms our belief and ignoring all the evidence that refutes it.

Some examples of confirmation bias are a bit more complicated than others and people typically mix together several different fallacious forms of argument. Here's an example from Denis Alexander's book Creation or Evolution (p. 213) that combines begging the question and confirmation bias.
Certainly God's great work in creation is distinct from his wonderful plan of redemption but the point here is that the whole book [the Bible] is God's book. The various narratives running through the book all stem from the same author. The evolutionary creationist1 is one who has a very firm belief in God's sovereignty over the whole created order, worked out in his plan and purposes for both creation and redemption. [my emphasis, LAM]

Once we have baptised evolution into our Christian worldview, just as countless Christians have been doing ever since the time of Darwin, then of course we will see no need to view evolutionary theory as the sinister intruder or 'universal acid' that the Dennetts and Dawkins of this world appear to imagine, but rather as the process that God has chosen by his sovereign will, to bring into being all the amazing biological diversity that we see all around us.
Denis Alexander assumes that the Bible is the inspired word of God and he spends a great deal of time explaining how the stories of Genesis actually conform to what science has discovered in the past 150 years. For example, Adam and Eve are not the first humans but the first couple to see God. The deluge was probably a local event. The opening verses of Genesis are allegory meant to convey an impression of God the creator rather than scientific fact.

This is an example of confirmation bias. Given that God exists and that the Bible is his word, then you can pick and choose those bits that seem to accord with science and this confirms your assumption.

But it's also begging the question since the real question is not why God was so ambiguous when writing his book but whether God even wrote it—or whether he even exists. That's the real conflict between science and religion but Denis Alexander ignores that question.

A few paragraphs later he brings up the idea of confirmation bias in a different context.
I am not suggesting for a moment that adopting the position of evolutionary creationism resolves at a stroke all the problems, and we shall be considering some of those in the next few chapters, but I would suggest that it provides a well-justified framework for continuing to hold together the book of God's Word and the book of God's works in a way that does justice to both.

Once we adopt this position, the ruse of annexing 'evolution' in support of ideological agendas loses its power to persuade. For example, there is nothing I can see in evolutionary theory that supports atheism. Of course, if we view evolution through an atheistic lens, we shall inevitably interpret it within an atheistic framework, as Dawkins does when he writes that in evolution he sees 'no design, no purpose, no evil and no good, nothing but blind pitiless indifference.' How could it be otherwise? The conclusions are built into the starting presupposition. This is what the atheistic worldview delivers; it is not what evolution itself delivers.
I don't think Denis Alexander sees the irony in his criticism. He criticizes Dawkins for confirmation bias but his crime is much greater and much more obvious.

If you were to believe Denis Alexander's description of atheism then there might be a kernel of truth in his criticism. But here's the problem. The question before us is whether science reveals a world full of purpose and design as the belief in Alexander's God requires. That's a scientific question. Dawkins says that the evidence does not support such a claim and I agree. (Alexander thinks that atheism is the metaphysical denial the gods exist—that's part of his problem.)

Thus, scientists such as Richard Dawkins see no reason to believe in creator gods. Neither do I. We don't initially start with the presupposition that gods exist nor do we start with the supposition that they don't exist. We are looking for evidence one way or the other.

The default position is that gods don't exist. That's the only way you can fairly examine the evidence. It's not confirmation bias, in my opinion, when the evidence doesn't support the claim of existence. The burden of proof is on the believers who are making an extraordinary claim.

Look at it this way. Imagine that you find no evidence whatsoever to support the existence of Santa Claus or the tooth fairly therefore you don't believe in them. Is that confirmation bias because you started out being skeptical of their existence? I don't think so.

We can argue about whether the universe shows evidence of purpose but that's a different issue. If it turns out that there's abundant evidence of purpose and scientists are ignoring it in favor of other evidence for lack of purpose, then that's confirmation bias (if they are atheists). But in this case, you can only accuse Dawkins of confirmation bias once you've established that he is biased in selecting his evidence. That requires a discussion about purpose but Denis Alexander doesn't go there in this chapter. You have to wait until page 424.

When you get to that page you'll discover that his arguments for purpose are very weak2 so he has no case when he accuses Dawkins of confirmation bias. There is a case, on the other hand, for accusing Denis Alexander of seeing only what he wants to see.


1.Denis Alexander describes himself as an evolutionary creationist.

2. He thinks the tape will replay in the same way and the evidence of convergence (Simon Conway Morris) proves it.

24 comments :

  1. Professor Moran, may i ask at at what age you became an atheist and what was the deciding factor? I'm not sure anyone can truly approach the subject from a neutral position, i think we all venture into our worldview with some preloaded bias. From that point on i think it's natural to defend what we believe to be correct. It puzzles me that atheists often blame indoctrination, geographical location, etc. on people being religious, then ignore the fact that they too are often products of their environment. Do you really believe that landscape can totally be remade to rid the world of religion?

    ReplyDelete
    Replies
    1. I was never a believer.

      It's silly to say that the lack of indoctrination is itself an indoctrination. You probably don't make such a claim for any other belief.

      Do you think that lack of belief in bigfoot is a serious problem that stems from the environment in which you are raised? What about lack of belief in leprechauns? Are you worried that this might also be a product of your environment leading to intrinsic bias?

      We have plenty of evidence that believing in gods requires indoctrination but there's no evidence that failure to believe in gods is a bias.

      Delete
    2. "i think we all venture into our worldview with some preloaded bias."

      Most atheists start with a preloaded bias to be believers; we often become atheists with a great deal of reluctance after we find religious belief to be supported by such poor reasoning and lack of evidence.

      Delete
  2. Larry.. I'm not sure I'd classify confirmation bias as a logical fallacy. It's certainly an error in thinking and it's one of the many biases that people often succumb to but a logical fallacy is something else. It is the use of faulty reasoning within an argument.

    Begging the question is a logical fallacy. Confirmation bias is a bias.

    ReplyDelete
  3. The title and the first word of the blog post say "conformation" rather than "confirmation".

    ReplyDelete
  4. Psychologists do refer to it as a cognitive bias.
    https://en.wikipedia.org/wiki/Cognitive_bias

    I guess you could lump cognitive biases and logical fallacies into a category of "things that fool you into believing false things"

    ReplyDelete
  5. Larry,

    Both your approach-"that gods don't exist-and Alexander's-"that God does exist"- are confirmation bias

    The only unbiased approach is the Socratic method by "following the argument wherever it leads".

    Take for example Albert Einstein's argument for "reason incarnate-regularities in nature that are mathematically precise, universal, and uniform.

    How could nature come wrapped up in such a sophisticated way?

    The unbiased argument leads to only one conclusion-an intelligent first cause.

    ReplyDelete
    Replies
    1. Free critical thinking lesson for you, Eric.

      If there was a debate over the existence of unicorns, and the pro-unicorn debater cannot provide any evidence for the existence of unicorns, the anti-unicorn debater wins. He doesn't have to even open his mouth.

      HTH.

      Delete
    2. Being surprised that nature is described by math is like being surprised that plants are described by botany.

      Delete
  6. You say "The default position is that gods don't exist." That sounds plausible because "gods" suggests something like the Greek pantheon, whose intrinsic probability is low. But if you rephrase it as "The default position is that the universe did not have a non-material cause" you might get more of an argument.

    ReplyDelete
    Replies
    1. That sounds plausible because "gods" suggests something like the Greek pantheon, whose intrinsic probability is low.

      How is that any lower than, say, the probability of the god of the Abrahamic religions?

      Delete
    2. " But if you rephrase it as "The default position is that the universe did not have a non-material cause" you might get more of an argument."

      Yeah, for example whether non-temporal beginnings even require causes. Whether there even was a beginning. And whether the non-material (Whatever that is) is anything like deities. It's unwarranted assumptions all the way down.

      Delete
  7. There's a very good chance that the story Noah's flood did derive from an oral history of a real event. Distorted oral histories appear to be a common source of myths world wide. (Of course, the story of Noah's flood came second or third hand from the earlier story of Gilgamesh.)
    The fact that parts of the Bible probably have some basis in actual events is no proof whatsoever that Yahweh or any other supernatural being actually exists. English boarding schools exist, too. That doesn't make Harry Potter a real person.

    ReplyDelete
  8. Non material cause of material?
    I think physics has a better handle on how to ask these questions, even if they are unanswerable.

    ReplyDelete
  9. The argument from Denis Alexander seems to be:
    "If there is a God then .... then something true is true"
    and he concludes "there is a God". That is fallacious.

    We would like to argue:
    "If there is a God then ... then something true is false"
    and conclude "there is no God", but the definition of God keeps getting changed by the other side until "there is a God" becomes "true". "true is true" is correct, but a waste of time. Asking for a hard and fast definition at the start doesn't seem to work.

    ReplyDelete
  10. “Alexander assumes that the Bible is the inspired word of God and he spends a great deal of time explaining how the stories of Genesis actually conform to what science has discovered in the past 150 years. For example, Adam and Eve are not the first humans but the first couple to see God. The deluge was probably a local event. The opening verses of Genesis are allegory meant to convey an impression of God the creator rather than scientific fact. ”

    These would be very odd positions for a supposed evangelical. Alexander’s theology is definitely cracked if that is his take on Adam and Eve. And his view on the flood does not square with the science that notices the planet-wide distribution of an estimated 50 million cubic miles of fossil-bearing sediments.

    ReplyDelete
    Replies
    1. @txpiper

      I was an evangelical like Alexander. Not all evangelicals are brainwashed fundamentalists who choose to reject science in favour of the bible.

      Some of them are accomodationists that choose not to reject anything specifically ruled out by science and so find ways to fit their beliefs about the bible to the scientific notions that they cannot reject.

      Off the top of my head, well known examples of this include:

      - Francis Collins
      - Dennis Venema
      - Peter Enns
      etc.

      Delete
    2. "...the science that notices the planet-wide distribution of an estimated 50 million cubic miles of fossil-bearing sediments."

      Huh? Are you claiming that fossil-bearing (fossiliferous) sediments are evenly distributed planet-wide? Are you claiming that all fossiliferous sediments are alike, including in the fossils they bear? Are you claiming that all sediments are fossiliferous?

      Delete
    3. The more txpiper writes, the stupider he reveals himself to be. Right, all those sediments could only have been deposited all at once, in a single world wide flood. Couldn't possible have been laid down in separate, local events.

      Learn to think, txpiper. You'll find it a useful skill.

      Delete
  11. I think evolutionary creationists are probably smarter when it comes to biology but dumber when it comes to theology than regular creationists. Do they not see all the horrible pain and suffering that occurred during evolution? Thats how God chose to create biological diversity?

    ReplyDelete
    Replies
    1. Meh... it's just as dumb to believe that all that horrible pain and suffering came about as a result of a pair of disobedient humans. It makes no sense either way.

      Delete
  12. Begging the question wasn't meant to mean that. Here the term is used as the equivalent of dodging a question and not as the logical fallacy of assuming the conclusion of an argument — a type of circular reasoning. It's still a flawed approach though and the text is pretty good.

    ReplyDelete