Knowing the Problem of Induction

Why Science and Religion are Incompatible, Part 4761

Once you eliminate the impossible whatever remains, however improbable, must be the truth (or a close approximation thereof). Almost A.C. Doyle

I have a friend who has often told me that as an atheist I rest too much on my preconceptions that God doesn’t exist for me to be open to evidence that his God does, in fact, exist.  He has told me that because of the problem of induction,  there is no way that I can “know” that God doesn’t exist, and that nothing in the scientific method can be used to support atheism.  Since no one can be justified, apparently, in drawing absolute answers from repeated observations, then it is silly to say that there is no God just because I have never experienced “Him.”

It’s the problem of induction, again.  I hesitate to discuss such a philosophical quandary among those who read this blog regularly; those who will likely school me on where my lack of formal philosophical training has failed me, but I have been thinking about the differences between science and religion as “ways of knowing.”

In order to maintain confidence that a causal relationship between natural phenomena has been established, one scientific method that I learned was to disprove a null hypothesis using statistical tools to analyze my data.  If the null hypothesis is not disproved, that means that the proposed hypothesis probably establishes a causal relationship and my investigation has yielded a good answer within a specified confidence interval. In other words, by following a scientific process, an investigator has come up with a good explanation for why something is so, or how something works.

This is only one of the methods that scientists use to discover how things work, one of the ways that people discover “how the world goes.”

Religion promises knowledge based on non-verifiable acceptance of authority, resignation to “mystery,” and the record of inscripturation.  Apologists for religion promise to provide “other ways of knowing” that aren’t limited to verifiable, positivistic methods. Religion, in general, tells people that we can know for certain that the supernatural exists and interacts in measurable ways with the natural.  Religion explains, in its “way,” the creation, miracles, interventions in personal lives and through catastrophic natural events.  The explanations are authoritative but not testable nor replicable through any reliable means.

There is a difference between the process of science and the nature of religion.  Science provides the “probable” answers, while religion promise certainty as long as the seeker will accept Mystery. 1 Cosmologists have teased out most of the probable answers as to what happened following the Big Bang to within Planck Time and are still trying to determine how this universe came into being.  They don’t know absolutely if the current understanding of the process of expansion has been accurately described, but they have reason to acknowledge that it has been described very accurately using the process of inductive reasoning.

Inductive reasoning, as I understand it, is the process of analyzing subsets of the whole to make rational judgments of the nature of the whole.  For a common example of how inductive reasoning works, I will use political polling.  A sample of the population of likely voters is queried as to how they plan to vote in an upcoming election.  The larger the sample polled, the more likely the pollster is to obtain an accurate prediction of the eventual outcome.  Once the sample size exceeds a certain level, the returns of accuracy and confidence change little and it would be foolish and expensive and time-consuming to sample more than necessary.  A poll of all the people who will vote would be the most accurate way to predict an election, it would yield an “absolutely true” result, provided that none of those polled were deceptive or changed their minds.

The ideal sample size can be determine through some quick calculations, for example:

Determining Sample Size

Determining Sample Size

((The trick to stats is designing the proper formula.  Once that has been done it is a simple matter of algebra.))

Nate Silver didn’t need to sample all of the voters in the 2008 election to predict that Obama would carry the electoral vote.  He merely needed to analyze the polls that sampled populations within the whole of the electorate.  The results he predicted were accurate to a specified confidence level, the famous “margin of error” of ±3 per cent.  There was a 5% chance that he could have predicted incorrectly.  In experimental design, a scientist will determine what margin of error will allow for the most probable and acceptable description of the causality of a natural phenomenon. Shorter:  Is this what caused that?  The  possible answers are not “yes” or “no.”  They are “probably” or “probably not,” or “that’s funny.”

The “problem of induction” is related to absolute knowledge.  If all knowledge is tentative, then any solution is as good as any other.  There is no certainty and there can be none, so my answer is as good as yours even if I haven’t done any serious investigation.  If you can’t state with a 0% margin of error that something is so, then you really have no useful knowledge. I can’t predict that the sun will rise tomorrow with absolute certainty, because I can’t see into the future.  I can confidently state that it will because I have an understanding that the sun doesn’t really rise, instead the earth rotates and creates an effective illusion that the sun is rising. For the Earth to stop rotating sometime in the middle of the night, events would be a bit more jarring due to the forces of momentum than I would care to deal with.  I wouldn’t then be too concerned that my prediction was wrong.

The “problem of induction” has been misused to claim that since there is no way to “know” that there is no God then God is likely to exist even if there is no direct nor indirect evidence of such an entity.  Not by any professional philosophers has this been done, mind you, but by friends of mine who think that they have stumbled onto something that “no atheist can answer.”

When it comes to the differences between religion as a “way of knowing” versus science as a way of understanding, religion offers something that science doesn’t.  Religion offers the comfort of absolute knowledge.  It offers the absolute answers, the answers that people want:  there is a creator that is watching after us and providing a way for us to experience a blissful afterlife.

The conflict between science and religion is in the means of acquiring knowledge.  Religious authority is often derived from personal revelations of prophets who have experienced something that to them is “real” and “true,” as true as the feeling of a burned hand in a fire. I have “felt” the presence of the Holy Spirit, but I have also “felt” the presence of the pagan Goddess in a drawing down of the moon. Both experiences were very emotional, uplifting, exciting and convincing. God’s presence was revealed to me, as was the Goddess’s. I should also note that both experiences were accompanied by prophecies from the respective supernatural agents.

Through these experiences, I found out how religious people “know” what they know. There could be no doubt, because the words came directly to me while I was experiencing the ecstasy. There was no induction needed, because through those experiences I had the Truth.  As Thomas Paine wrote in The Age of Reason:

It is a contradiction in terms and ideas, to call anything a revelation that comes to us at second-hand, either verbally or in writing. Revelation is necessarily limited to the first communication — after this, it is only an account of something which that person says was a revelation made to him; and though he may find himself obliged to believe it, it cannot be incumbent on me to believe it in the same manner; for it was not a revelation made to me, and I have only his word for it that it was made to him.

I could tell you the Truth of those prophecies, but you would have to take my word for it.

Religion and science are not compatible because of the illusory nature of “truth.”  We all have truths.  Religion claims to provide Truth.  Science is just a process that uses methods to get close to truth.  Religion provides other ways of knowing.  My question is in knowing what?  What does religion help us know, exactly?  And if induction can’t be used to prove an absolute, is that really a problem that religion can solve?

I don’t know.

  1. Have you ever noticed that Catholic theologians pronounced the word “mystery” with the “M” capitalized? How do they do that? []

Tags: , , , , , , ,

16 Responses to “Knowing the Problem of Induction”

  1. July 15th, 2010 at 6:07 am

    Lorax says:

    I would concede that as an atheist I cannot conclude definitively “there is no god.” However, I can conclude there is no this god or that god. What I hate with these arguments is that the believer believes in a specific god that did and does specific things, but when faced with a contrary position, the believer then resorts to a god concept that is immune to disproof but also cannot possibly be the god they believe in.

  2. July 15th, 2010 at 6:12 am

    Mike Haubrich says:

    It’s frustrating, isn’t it? Like a shifting of the goalposts each time to take a shot and think you have scored only to find that the target is no longer there.

    Deep down, I am an agnostic towards the existence of any particular God(s) and I call myself an atheist because as you have concluded there is no reason to accept the existence of any particular god.

  3. July 15th, 2010 at 8:50 am

    Alden says:

    “Religion” is, unfortunately, too broad of a term to use when discussing epistemology. The answer to your question, “Know what?” with respect to Christianity is simply, “God.” That’s it.

    The problem I see with most “people of no religion” is that, to oversimplify, is that religion is mostly round pegs, and that you only have square holes. The discussions between Christians and atheists often don’t even make sense.

    You might enjoy the book “Descartes Bones,” which does a fairly decent job of discussing the issue of faith and science.

  4. July 15th, 2010 at 10:19 am

    Mike Haubrich, FCD says:

    Yes, I am purposely being broad here. Christianity doesn’t get an exception for reliance on revelation, or Truth Claims. Another religion will say “Know Vishnu” where you say “Know God.” Or “Know Alllah.” This, however is conflating, because religion is saying that there are ways to “know” things that science can’t be used to discover. To “know” is a strong claim, and no religions have demonstrated a method of determining Truth with a capital “T” in an unassailable manner. Yours has done no better than any other.

    I have heard of “Descarte’s Bones” and intend to read it. Thanks for reminding me.

  5. July 15th, 2010 at 8:02 pm

    Boz says:

    ” The results he predicted were accurate to a specified confidence level, the famous “margin of error” of ±3 per cent. There was a 3% chance that he could have predicted incorrectly. ”


    A 2-party-preferred polling result for Candidate A of 51, with a margin of error of 3% means that:

    the True result (if we ran a population-wide census, with no liars and 100% response), has a 95% likelihood of being within the range 48 to 54.

  6. July 15th, 2010 at 10:23 pm

    Mike Haubrich says:

    I knew that, too, and will fix it.

  7. July 16th, 2010 at 8:14 pm

    FormerComposer says:

    To turn things on their head a little, “What is the question for which God is the answer?”

    In other words, applying Occam et al, starting with an answer before having a question seems somewhat ill-conceived.

  8. July 17th, 2010 at 2:37 am

    DiscoveredJoys says:

    To be even more picky religion and science are not different ways of knowing but are both ways of experiencing the feeling of knowledge.

    It’s just that science, done properly, is a method of confirming that your experience of feeling about an item of knowledge is similarly experienced by other people, and may be used to build up the shared experience about other facts.

  9. July 17th, 2010 at 6:09 am

    Mike Haubrich says:

    Yes! That is one paragraph to say what my full essay tried to say.

  10. July 18th, 2010 at 5:43 am

    cipher says:

    The problem I see with most “people of no religion” is that, to oversimplify, is that religion is mostly round pegs, and that you only have square holes.

    Of course – the fault is ours, for not having round holes to fit your pegs. The fact that you’re missing sqaure holes to fit ours seems not to have occurred to you.

    In the theists’ view, it’s always the atheist who is somehow “insufficient”.

    The discussions between Christians and atheists often don’t even make sense.

    Yes, I wonder why that is?

  11. July 19th, 2010 at 9:35 am

    Oedipus says:

    What does religion help us know, exactly?

    My view is that religion does contain some “truth” in the (albeit vague) scientific sense, but it’s not the truth the religion itself claims. Religion is ultimately a psychological experiment, with repeatable results. It is a high-level description of certain psychological phenomenon, couched in provincial terminology, having both positive and negative effects on the “patient”. Religion says something about the neural network inside our heads, and that is a scientific statement, however nondescript. (Though I realize you probably meant know with regard to the patients, not the observers.)

    Regarding induction, the classic example is finding confirming instances of the statement “All crows are black.” But that statement is equivalent to its contrapositive, “All non-black objects are not crows,” and there are infinitely many confirmations of that. Does a green parakeet confirm that “all crows are black”? A brown kangaroo? A white rock?

    Out of amazing coincidence, I happened to have recently mentioned the feeling of the Holy Spirit in similar light.

  12. July 19th, 2010 at 7:53 pm

    NewEnglandBob says:

    There is no truth in religion. It is nothing but flim-flam, deceit, obfuscation, lying and attempt to control others.

  13. July 19th, 2010 at 7:54 pm

    Deen says:

    Not by any professional philosophers has this been done, mind you

    I wouldn’t be so sure…

  14. July 19th, 2010 at 9:33 pm

    Oedipus says:

    Bob, if you were responding to me then granting your assertion my point would be that religion is the science of flim-flam, deceit, obfuscation, lying and attempt to control others.

  15. July 28th, 2010 at 3:47 pm

    OH says:

    I don’t agree that “if you can’t state with a 0% margin of error [i.e. with 100% certainty] that something is so [i.e. true], you have no useful knowledge”. From a scientific standpoint I see no use for 100% certainty or truth. We build models or theories that appear to be the best possible fit with what we observe, and we continuously test these, e.g. in the light of new findings or theories. I wouldn’t even call that “getting closer to the truth” since that implies that truth exists in a some absolute ultimate sense. As I see it, that also dissolves (makes irrelevant) the so-called problem of induction. Frankly I believe it is misguided to take ordinary words like truth and certainty from the everyday language where they originated and apply them to science or philosophy. It ensures a never-ending discussion for no good reason. As for truth in religion, if believers find comfort in equating belief and conviction to truth, I would not argue against it, precisely because it is belief or faith based.

  16. July 28th, 2010 at 7:35 pm

    Mike Haubrich says:

    And that is what I was driving at. I didn’t really note it as a facetious statement, hoping that someone would challenge me on it. And you certainly picked up on the point of the post.

SEO Powered by Platinum SEO from Techblissonline