Sunday, October 11, 2009

GRE study book on God's existance - failing at logical fallacies

So if you follow me on twitter at all, you have probably figured out that I've been cramming for the GRE (test to get into grad school). I'm taking the test tomorrow morning, so I've been reviewing one of those study guide books. I'm not too nervous, but I figured reviewing can't hurt, especially since I haven't done any math other than plugging in numbers for about three years (yay Purdue's science curriculum).

I think I'm pretty much golden on the writing section. The first part you have to be able to express an opinion, and the second part you have to analyze an argument and find errors in their reasoning. Yeah, I think I'm pretty good at the whole being opinionated and criticizing faulty reasoning. Regardless, I started reading the section of Logical Fallacies...and found this:

Shifting the Burden of Proof
It is incumbent on the writer to provide evidence or support for her position. To imply that a position is true merely because no one has disproved it is to shift the burden of proof to others.

Example: Since no one has been able to prove God's existence, there must not be a God.

There are two major weaknesses in this argument. First, the fact that God's existence has yet to be proven does not preclude any future proof of existence. Second, if there is a God, one would expect that his existence is independent of any proof by man.
Are you kidding me? This is in the section on how to think logically? My only gripe is that the sentence says "must" - I would lower it to "most likely" or "probably" because yes, absence of evidence is not evidence of absence. But regardless, the explanation they give is itself illogical. One, the burden of proof should lie on those who make claims. An absence of something extraordinary isn't a claim - it's a null hypothesis. If you have absolutely no proof, what is supporting your argument? Secondly, it's effectively impossible to prove a negative (that something doesn't exist), which again leaves the burden of proof with those making the claims. Three, future proof doesn't hold any ground in current arguments. If I said that I might potentially eventually have proof that unicorns exist, would anyone take me seriously? If I said one day scientists may find that a diet of nothing but chocolate is good for your health, should we all eat nothing but chocolate? No, because it's not real evidence. And finally, the last sentence about God's existence being independent of any proof of man is a logical fallacy I like to call "Making Shit Up." Why is God's existence independent of any proof of man? What reason do you have to think that other than conveniently and arbitrarily defining God that way? Why God and not gods, or goddesses, or aliens, or fairies?

Logical fallacies when trying to teach logical fallacies. Thank you, book.


  1. It's hard to argue and fight null hypotheses in this medium

  2. That, I think, is terrifying. It demonstrates the degree to which fallacious arguments are accepted as logical. The burden of proof rests on the POSITIVE claim; the idea that something or some effect DOES exist. The negative is assumed until proven otherwise. The worst part, however, I think you missed; the fact that this was in a GRE study guide will itself be used as a "proof" that arguing against gods are illogical, as if books are somehow beyond criticism.

  3. Ugh. That is, indeed, illogical, although it's framed in such a way as to make it difficult to express why. I think I see what they're trying to get at, though.

    The "precludes any future proof" bit is just a bad way of wording "absolute certainty of a negative claim is impossible", which is trivially true and not very interesting (I think this is why they included "must" in the proposition; including "must" does actually make the argument flawed, though not for the reasons they claim).

    The "independent of any proof by man" bit is rather bizarre, though. I think what they're trying to say is that an unobserved phenomenon can still exist despite our lack of knowledge thereof. This is also trivially true, yet it's no reason to assert an affirmative claim to the existence of such a thing; it's precisely Russell's teapot.

    Of course, there's still the problem that they're treating "God does not exist" as an AFFIRMATIVE claim and trying to use it to illustrate "shifting the burden of proof". Bloody hell no. They just committed the exact fallacy they wanted to illustrate. I find it rather sad that they had the perfect example right in front of them and managed to get it wrong.

    Hmm, so it seems they got two minor points correct and missed the major one? We have an expression for that: a stopped clock is still right twice a day.

  4. I’m scaring myself. When I first read that excerpt, I could tell something was iffy, but I couldn’t even put my finger on just what was wrong. I had to read your explanations to realize what was teasing my Skeptey Senses, which is unsettling for me, considering I believe I’m rather good at this sort of stuff.


  5. Hi When exhausted by the Is there a God or which religion is the true one?, go beyond all that and read a short book called BEYOND, The End Of Faith.
    Deals with concepts of how can we overcome this primitive quagmire and what would a mature society look like. You can download to a P.C., Kindle or other and by placing zeroes in the price box, it is free. Lynn Dewey

  6. I just took the GREs this past week and I have good (but frustrating) news and bad news...

    The probably good: Princeton review exams are way harder than the actual thing-- you can probably relax a little. The Powerprep software gives you a much better idea of where you are scorewise.

    The bad: If you log on to Princeton's webpage for tests, it may ask you if you want to find out if you have what it takes to be a naturopathic doctor.

    Princeton review was my hero when I was studying for AP. Not so much anymore.

  7. Of course, assuming that god has omnipotence (generally a typical characteristic of God) then the second statement is more than just bullshit.

    Any God worth his omnipotence is certainly able to make himself provable, such that if he cannot be proven, it is by his choice.

    Granted, not an issue for logic, but certainly a conundrum for any theist.

  8. Uhh... their second point is a bit baffling and difficult to defend from the "Purely logical" standpoint, which is how I would try and defend the author since it's just general GRE help. We can tell where the author of this little bit of GRE-help sits on the theist debate :D.

    I think the first part follows, though, in the strictly Aristotelian logical sense. It ignores a good chunk of epistemological gains outside of "pure logic" , but at least given Aristotle's thoughts on logic, I think it follows. (And perhaps even outside of Aristotelian logic... but I don't think it even follows in Kantian logic, and he is the next big contributor to logic to my knowledge. You need evidence of some kind to support a claim, simple as that. If you admit there is no evidence, then you have no claim worth considering). There's also something about the wording of the position that gives credence to the first point, I think: Specifically because the atheist position considered is stating "Must" it is a logical fallacy. That does not follow. Usually an atheist will say something more along the lines of "Given all the arguments I've seen for God's existence, I'm reasonably certain that God does not exist". That's a statement taking into consideration things outside of strict Aristotelian logic.

    Of course, Aristotle also thought rocks stopped moving because that was their "Natural" way of being -- and that was also consistent with his logic.

  9. Afraid I'm going along with the book. [There must not be a God] is the positive claim with the burden of proof being on the claimant.

    If the book is trying to use this to suggest this is what most atheists are claiming then it's a strawman. But then books about logical fallacies often put examples people wouldn't make to make a point.

    To consider this a logical book fail you would have to consider the statement [God must not exist] NOT to be a claim. Of course [must] is the word that changes everything but that's exactly what logical arguments are about: it could rest on a single word.

    At worst it's an unrepresentative example because virtually no atheists make the claim and someone reading it could get the wrong impression. But if it was about unicorns it would be correct (just as Russell was correct that we can't know that the teapot MUST not exist).

  10. Well I think you've just figured out what to do on the second part of the written GRE...

  11. Premise: No one has been able to prove God's existence.
    Conclusion: There must not be a God.

    I think we can agree that the statement is fallacious but not for the reason that the book states? The second part about how "his existence is independent of any proof" just mystifies me though.

    By the way, good luck.

  12. If absence of evidence is not evidence of absence, then what are you looking for when you look both ways crossing the street?

    Absence of evidence is not PROOF of absence, but it can be evidence of it. Especially when you would expect evidence for a particular phenomenon.

  13. Echoing Jeff, Absence of Evidence is Evidence of Absence. In fact, in Bayesian terms of belief, it's mathematically required.

  14. Well, it could be worse. This could be the fill-in-the-blank/multiple guess question you mentioned on Twitter... at least this one can potentially be permuted into a reasonable form if you read it the right way.

    That other question... I'd have chosen scorn as well, I think, it's the only one that makes any sense. As far as "lose" goes... urgh, utter nonsense - except perhaps in the pedantic sense in which one cannot lose something one never had? I don't think I can stretch it far enough to convince myself that's what they meant...

    Good luck with the test. Let's hope these questions aren't representative...

  15. I'm having a protracted (and rather frustrating) online conversation with what I would consider a serious theologian. He has made the same remark about not expecting evidence for the existence of god, which I find baffling, considering the attitude and characteristics attributed to the christian god.

    OTOH he also constantly does that irksome shifting the burden of proof thing (more like sending the burden of proof into outer space thing) so I'm not that surprised...

  16. ugh, my head hurts

    good luck on your test!

  17. Oh, and while the distinction I'm making is snooty and hoity-toity, I think it's important because my impression is that "Logical" often can be subbed in for "Logically Aristotelian". This is, of course, just a personal impression, so take it with salt, but it kind of runs along the lines of the validity/soundness distinction, and might be worthwhile to brush up on for test taking purposes.

  18. Premise: No one has been able to prove God's existence.
    Conclusion: There must not be a God.

    It was an editorial error. There was meant to be TWO premises.

    Best of luck with the exams!

  19. That's some outrageous bullshit right there. A workbook like that should only be trafficking in unambiguous, uncontroversial answers. Even a huge swath of religious theists would reject the idea that God cannot be proved through reason and evidence. It's not even just prejudiced in favor of theism but in favor bald, baseless irrationalism.

  20. My personal favorite, when prepping for the GRE general was when the textbook described 'Slippery Slope' arguments as a fallacy, and then presented a top scored paper which, under its second last paragraph started with "A slippery slope may follow too...".

    Or there was the questions which didn't give enough information to come up with the conclusion. McGraw-Hill failed pretty epically on that one.

    Still, the GRE-General test is pretty straightforward. Good luck.

  21. Actually I think their arguments are sound.

    From a logical standpoint, having no evidence or being unable to prove something doesn't negate anything's existence. The best we can say about unicorns is: "we have not found evidence for the existence of unicorns." You can never say "unicorns don't exist", categorically. Not in the same way you can say "1 + 1 does not equal 3". The best logical thing you could say is "because of the complete lack of evidence, the likelihood of unicorns existing is vanishingly small".

    On the second "weakness" they explore is defensible as well. If you define something as "godlike" in a traditional sense, one of its properties is almost certainly to be a transcendence of man or to have preexisted before man. It is plausible that if you hypothesized the existence of a god, that you would not be able to prove – with your human brain, which is inferior - his existence.

    Another way of looking at it is: it is impossible prove something that is defined to be unprovable. Logically, this is completely fine. The trick is accepting the property of unprovability. Most of us would agree that starting off with a definition of something as unprovable is unsatisfactory and wouldn't even go there. However, many religious people feel it's a fine and dandy place to start. This statement they make can only really logically flow if you first hypothesize the existence of a transcendent god, which is a huge stretch in and of itself.

    Logic doesn't really have to do with evidence, it has to do with analysis of statements and their consistency with one another. The example those guys gave is perfectly correct from a logical standpoint, but it is colored in a way that upsets an atheist because it doesn't explore the much stronger arguments about God's non-existence (the vanishingly small tangible evidence, the biological disposition towards belief, the anthropologic development of religion, etc) . I think it's a bit of a throwaway example of logical analysis that is in poor taste, however accurate it might be.

    The key point they want to make is that shifting the burden of proof is not a satisfactory defense of a position. They certainly got you to think about it, didn't they? :)


  22. If you define something as "godlike" in a traditional sense [than] you would not be able to prove – with your human brain, which is inferior - his existence.

    With the premise the proofs are only good for math, I'm not really convinced about this. The actions traditionally attributed to gods are things like lay waste to continents, flood planets, create universes, answer prayers... you'd expect to se at least some evidence.
    The logic of the proposition is not that sound, unless you take e.g. the bible and strip away all the interventions from god for which we would expect to find evindence (and end up with a three pages leaflet and a deist position:p)

  23. Good information about GRE it is very useful for students

    Study Abroad