Comments

  1. mond says

    I am so disappointed in Matt.
    I thought that the quality of hero worshipper he would attract would be of a much higher quality.
    That guy was full of so much new age type woo that it was painful.

    As a result I have had to relegate Matt from hero status to some guy on the TV/Internet who sometimes say interesting stuff.

  2. mond says

    @#2
    This was an attempt at some self deprecation meta humour.
    The idea that Matt is no longer my hero because some else hero worshipping him did not meet my high standards is obviously ludicrous.

    Thanks for making me explain the Joke…you’re a bad person.

  3. Chris says

    Three arguments were brought up that I think merit a more rigorous refutation; fine-tuning, consciousness, and the Kalam cosmological. It would be helpful to provide counter-arguments to the strongest versions of the apologetics, even if the callers themselves aren’t familiar with strongest versions. I hope the hosts take these into consideration so that they can be better prepared to discuss these topics on future shows.
    .
    Fine-tuning and consciousness both commit the fallacy of understated evidence:

    Paul Draper has usefully identified a fallacy of inductive reasoning he calls the “fallacy of understated evidence.” According to Draper, in the context of arguments for theism and against naturalism, proponents of a theistic argument are guilty of this fallacy if they “successfully identify some general fact F about a topic X that is antecedently more likely on theism than on naturalism, but ignore other more specific facts about X, facts that, given F, are more likely on naturalism than on theism.”

    .
    On fine-tuning:

    1. Our universe is not teeming with life, including life much more impressive than human life.
    2. The only intelligent life we know of is human and it exists in this universe.
    3. Intelligent life is the result of evolution.

    Further reading: Hostility of the Universe to Life: Understated Evidence about Cosmic Fine-Tuning?
    .
    On consciousness:

    1. Conscious states in general are dependent on the brain.
    2. The very integrity of our personalities are dependent on the brain.
    3. The apparent unity of the self is dependent on the brain.

    .
    On the Kalam:
    Decisive Refutation of the Kalam Argument
    Excerpt from the author Wes Morristion:

    When I try to conjure up a picture of something — a house, say — “popping
    into existence” without a cause, it does seem pretty absurd. It may
    not be logically impossible, but it is inconsistent with everything I know of
    the world in which I live. Houses don’t just materialize out of nothing.
    They have to be built.
    So far, so good. But I don’t see how considerations of this sort can he
    appealed to by someone who believes in creation ex nihilo. After all, a
    house “popping into existence out of nowhere” doesn’t seem any less
    absurd just because somebody says (or thinks), “Let there be a house
    where there was no house.”

    .
    Final comment on the caller who offered an objection to fine-tuning: Although he had some trouble articulating his argument, it did touch on what I think is a strong point to make.
    Theism doesn’t actually predict physical bodies. Rather, it predicts minds or moral agency. God could make it the case that his moral theater is played out entirely in the immaterial realm.
    Given that intelligent life of some sort exists, observations of FT for physical life strongly support Naturalism over Theism because there are many more ways for intelligent life of some sort to exist without fine-tuning on Theism. (Life could be sustained with perpetual miracles, God would not be restricted to certain values of physical constants) Given that intelligent life of some sort exists, FT is the only way to bring it about on Naturalism. If we assume theism is true, it’s an incredible coincidence that intelligent life of some sort came to exist on the only option available to Naturalism.

  4. Hippycow says

    Yeah, I didn’t quite get that. “You’re a personal hero of mine.” Yet apparently doesn’t understand or agree with anything Matt says about evidence-based reasoning, or skepticism. So, Matt’s his personal hero because maybe he has a nice tone of voice? Was it the pink beard? Maybe he just has Matt confused with Deepak Chopra? That’s easy to do and I’m sure it happens all the time.

    Maybe Oz was Dr. Oz? Or was it Oss? Maybe that’s spelled A-S-S? Anyway, dude, you seemed to have a little confusion about how to choose your personal heroes. Maybe you can take a homeopathic remedy or get some acupuncture to help with that problem.

    By the way, John, I loved the quip about being skeptical of astrology because you’re a Sagittarius and they’re skeptical. I’m still chuckling. I can’t wait to use that one next time the opportunity arises.

  5. Rick Pikul says

    A suggestion that came to mind while watching the aftershow: In the open threads, include any links brought up during the show as part of the post. That way there is never any issue of getting spelling across or having people miss the URL because they can always come here to get it.

  6. Hippycow says

    Additionally, @Chris #4: I really like your thinking in your final comments on fine tuning; these points have often occurred to me as well. Why does god need a temporary physical stage for humans to be tested before they are dispatched to eternity in heaven or hell? Why create a universe with particular static characteristics for this stage, when you could maintain it by pure will with no particular consistent rules? A god could really create a flat earth without gravity, with firmament above, etc. and maintain it all magically. There would be no need to have cohesive rules.

    By the way, it is also worth noting that the fine tuning argument is often over stating the evidence for fine tuning, in that many of the “constants” we have are dependent on other constants and can really boil down to just a few constants. I think the jury is still out on this stuff, but somehow I doubt a “professional philosopher” of the WLC stripe is going to be among those doing any breakthrough discovery in this department.

  7. Muz says

    Since the audio came up I can wade in without fear of seeming too obsessed (OK that probably won’t work, but anyway…)

    I noticed particularly that the last three weeks the MP3 and the video (where available) were creamy deliciousness . Well that might be overdoing it but they were fine. Little to no distortion not introduced by phone systems and so forth. But the stream was awful. Hence the sourness of chat and possible confusion of whoever is on the desk.
    So the stream goes to its own separate output maybe? Ustream hates your broadcast levels for some reason? I’m obviously guessing wildly I hope it is some rarely touched pot or whatever computer handles the job so it can be easily adjusted without messing up the flow. Otherwise since every other format seems a-ok I guess leaving it might be the better option and we streamers will just have to suffer a bit..

  8. says

    “I hope the hosts take these into consideration so that they can be better prepared to discuss these topics on future shows.”

    There’s no doubt that I was “off” on Sunday. Difficulty with the fire triangle made that glaringly obvious. That said…

    “Fine-tuning and consciousness both commit the fallacy of understated evidence…”

    Draper’s proposed fallacy is, to my mind, better identified as “cherry picking” and the baggage he adds to it isn’t very good. He seems to make the same error that I identified on Sunday: comparing likelihoods where one is unknown. As here: “more likely on theism than on naturalism […] more likely on naturalism than on theism.” There seem to be many hidden assumptions being made about the specifics of what theism entails or permits.

    You seem to expand on this error in your notes on fine-tuning…implying that there’s some way to calculate or approximate which is more likely when one of the two proposed options (as far as I can tell) can’t be quantified or approximated.

    Over the course of 10 years on the show, I’ve answered the fine-tuning argument many different ways, including making remarks similar to yours. I intentionally try different approaches so that I can see which sort of responses are actually more effective, rather than which ones I merely prefer (and so that the show doesn’t become a scripted response). Sometimes it’s effective, sometimes it isn’t. It’s a mistake to assess a single episode – but it’s good to get the feedback to find out what we should avoid.

    That said, I don’t find your response (here) to the fine-tuning argument to be a very good one, for these reasons:

    Point 1

    “1. Our universe is not teeming with life, including life much more impressive than human life.”

    This raises good points that don’t really address the fine-tuning argument. The argument doesn’t claim that the universe is fine-tuned to be teeming with life, it merely argues that the universe is fine-tuned to permit life (specifically life-like-us) and that theism is the best explanation for why this is permissible. The abundance of life is irrelevant. The fine-tuning argument also doesn’t raise issues about more impressive life, so that seems to miss the mark, as well.

    The point of contention is their claim that theism is a better (more probable) explanation and my response requires them to demonstrate how they determined this. Any response that attempts to argue that naturalism is actually more probable must buy into their hidden assumption that the probability of theism is somehow known.

    It’s not, is it?

    Point 2

    “2. The only intelligent life we know of is human and it exists in this universe.”

    This isn’t really a response to the fine-tuning argument, either and it spawns additional problems by claiming that the only intelligent life we know of is human. The fine-tuning argument doesn’t specifically address intelligent life, and depending on the definition of “intelligence”, humans may not be the only example. Meanwhile, “that we know of” is incredibly problematic. Once upon a time the only swans we knew of were white.

    Point 3
    “3. Intelligent life is the result of evolution.”

    This doesn’t address the fine-tuning argument and not merely because it’s talking about intelligent life.

    It implies that evolution is incompatible with theism. I agree that this is, in fact, the case (in fairly narrow definitions), but there’s a lot of work required before this can be of use. It opens up discussions about abiogenesis and whether or not a god can work within the process. Again, I understand the problems here…but the statement, on its own, could be viewed as an unsupported assertion (there is a dead link on the blog that you linked to for this, so I couldn’t track back to the thinking behind it).

    It also implies, in black swan fashion, that this is the ‘only’ or ‘most likely’ way to arrive at intelligent life based merely on what we have observed – with no demonstration that this is the case. It’s as empty as saying “A straight flush arises from a shuffled and fairly dealt deck of cards”…it’s true, but it can also arise from artifice. (Which of the two is more likely? There may not be sufficient information to calculate which is more likely.)

    “On consciousness:
    1. Conscious states in general are dependent on the brain.
    2. The very integrity of our personalities are dependent on the brain.
    3. The apparent unity of the self is dependent on the brain.”

    I’m pretty sure that I made those points, in some form. Though I’ve answered this question several times in the past week and may be conflating different responses.

    Suffice it to say that I’m in agreement here.

    “On the Kalam:”

    There are a number of different responses and I’ve offered a number, myself. I’m not sure that we really spent much time on the Kalam on Sunday…but it’s worth noting that giving a 5-10 minute, in-depth, philosophical response to a loosely presented version of an argument isn’t, in my opinion, a good idea for most scenarios on the show.

    I’ll be happy to do it in a formal debate, and I have.
    I’ll be happy to do it in an instructive format, like my Patreon project (and that specific argument is on the ‘to do’ list).

    But the show is a conversational exercise that has to be tailored to the understanding of the caller and approachable by almost any audience.

    I’m pretty confident, though, that if you look back on other episodes that addressed the Kalam, you’ll find them more to your liking.

    Finally:

    “Theism doesn’t actually predict physical bodies.”

    So what?

    It’s magic. It could be argued that it predicts either nothing or nearly anything.

    “God could make it the case that his moral theater is played out entirely in the immaterial realm.”

    Or, it might predict that the physical realm is necessary for certain moral understandings…or, in fact, for morality – at all.

    What we tend to identify as morality is specifically dependent on physical human interactions in a physical world.

    While religions, Christianity in particular, often make claims of immorality that are based on thought crimes or crimes against a god (impiety, apostasy) the bulk of the instructive laws address our interactions in this world (stealing, lying, killing, maiming, etc.). Those make no sense outside the context of a physical existence.

  9. Esquilax says

    The real problem with the fine tuning argument is that it assumes its conclusion: how did the arguer determine that the universe was, in fact, “fine tuned” for life at all? Fine tuning, the whole reason that’s an argument for theism, is that it requires a fine tuner, but nowhere in the claim that the universe contains highly specific constants that are conducive to life is there an argument for the assertion that those constants must have come about by the intelligent manipulation of an outside agent; it is equally possible, under the premises of the fine tuning argument, that those specific constants rolled out the way they did naturally, because at no point is there even an effort to establish that these specific constants are some sort of goal that were being consciously worked towards, rather than just the outcome we ended up with. An unjust level of significance is placed in the fact that life could arise here, but at no point is there an attempt to justify that.

    It’s really a non-starter until the theist can first establish that there is, indeed, fine tuning… which is also the conclusion they’re trying to get to. It’s basically like saying nothing at all.

  10. Chris says

    Reply to comment 9, which is a reply to comment 4.
    .
    Thanks for the reply, I think we are both on the same page that I offer these comments in the spirit of identifying strong arguments against theistic claims and my hope is to provide useful feedback.
    .
    A few points of clarification; I think you did touch upon some of the points I made on the show, I put together my comment with the intention of being comprehensive and to frame the points within the context of the fallacy of understated evidence. I would also like to say that I understand you are working within the format of the show, and that you frame your answers with many considerations in mind.
    .
    I think two points should be emphasized:
    ~These arguments apply to general theism, since Christianity entails theism while making additional assertions, an argument against theism is an argument against Christian theism.
    ~These are inductive arguments, with the goal of assessing the final probability of a given hypothesis.
    .
    On comparing theism and naturalism Jeffrey Lowder writes:

    It is a commonplace in confirmation theory to measure the ratio of one explanatory hypothesis to another logically incompatible explanatory hypothesis, even if those two explanatory hypotheses are not jointly exhaustive. Again, even if the two explanatory hypotheses are not jointly exhaustive, demonstrating that the ratio of the probability of one hypothesis to the probability of another is greater than one can be used as part of a larger argument to show that the former hypothesis has a high final probability.

    He provides a more detailed overview here.
    .
    On the Fallacy of Understated Evidence:
    I think you’re dismissing the fallacy too quickly. As an example, let’s consider the case of a murder and “John” is a suspect. If it’s established that John was in the city of the murder on the night it occurred, then that is a general fact that supports the hypothesis that John is guilty. However, if it were the case that John was at a basketball game at the time of the murder, using the general fact that John was in the city to support the hypothesis that John is guilty would understate the evidence that John was not at the scene of the crime.
    .
    Ultimately I think you’re underestimating the field of inductive reasoning, probabilistic arguments are less ambitious than deductive arguments but are also harder to counteract. By dismissing these arguments and arguments like them, you’re cutting yourself off from a valuable tool in determining the epistemic strength of a given hypothesis.

  11. Hippycow says

    Anyway, don’t constants just come about as an artifact of our units of measure?

  12. EnlightenmentLiberal says

    Anyway, don’t constants just come about as an artifact of our units of measure?

    To some degree, yes. However, there is a unit-less constant that relates the mass of the proton to the mass of the electron, and it’s the same number no matter what units you use (because it’s a unit-less constant).
    http://en.wikipedia.org/wiki/Proton-to-electron_mass_ratio

    AFAIK, there are other such constants too.

  13. says

    “By dismissing these arguments and arguments like them, you’re cutting yourself off from a valuable tool in determining the epistemic strength of a given hypothesis.”

    I disagree…and appealing to incalculable probabilities isn’t really getting anyone anywhere.

    And pointing me to Lowder doesn’t really help either. It’s not like I’m going to agree based on who says something.

  14. EnlightenmentLiberal says

    Can’t wait to watch the show. Should be interesting based on the comments here.

    @Matt and Chris
    I have found that Matt isn’t sufficiently Bayesian sometimes IMHO. (If you’re reading this Matt, I strongly suggest the book “Proving History” by Richard Carrier.)

    I find much to like about post 13, but I don’t have a broad enough context to know if that’s a rebuttal to some position of Matt, or just some tangent or non-sequitir.

    However, I’m also strongly with Matt when he says that fine tuning arguments are generally bunk because we have no idea how you might approximate the relative probabilities of an intelligently designed universe vs one that is not intelligently designed.

    Now, I recently had some related conversations in person with Richard Carrier (I’m just a fanboy). Richard argues IMHO rightly that as a first approximation, simpler theories are more likely to be true than more complex theories. It’s a basic restatement of Occam’s Razor. Richard also argues that we can put numbers to the complexity of the general natural hypothesis and the general supernatural hypothesis and thus deduce a relative probability, and I balk at that.

    Richard defines supernatural as any irreducibly mental substance – whatever that means. I think the definition is logically inconsistent and incoherent because the concept of “mind” is properly defined in terms of some sort of reduction. (For that point, please see the work of Daniel Dennett.) Thus IMHO Richard’s entire argument breaks down.

    Perhaps if someone else gave me proper and workable definitions of “natural” and “supernatural”, we might be able to calculate complexities of the hypothesis specification, and thus deduce a relative probability. However, I think the words “natural” and “supernatural” are quirks of English which confuse us into making wrong conclusions.

    Funnily enough, all of this relates to an ongoing disagreement between Matt and I (mostly one-sided because I’m a nobody). While Richard at least has a specific definition of supernatural, I don’t think I’ve yet seen Matt even give a definition of the term, but Matt readily says that one cannot use science on supernatural – a claim which is either wrong or vacuous. To use Matt’s own example against him, the following hypothesis is totally testable: there is a creature who calls himself Thor and who can apparently conjure long-lasting localized lightning bolts and throw them. Perhaps this hypothesis is not a supernatural hypothesis, but then I argue that by this standard, there is no such thing as a supernatural hypothesis with observable predictions. Thus, Matt’s claim about intrinsic methodological naturalism (a term used by the paper below) is either wrong, or vacuous and thus wrong-headed.

    For additional information, I would strongly suggest Scott Clifton’s Skepticon 7 talk and the following academic paper.

    How not to attack Intelligent Design Creationism: Philosophical misconceptions about Methodological Naturalism

    Maarten Boudry, Stefaan Blancke, Johan Braeckman

    https://sites.google.com/site/maartenboudry/teksten-1/methodological-naturalism

  15. Chris says

    In reply to comment 16
    .
    The quote by Lowder was intended to provide clarity and the article was intended to provide reasoning that stands on it’s own independent of the person making it.
    .
    I think the main point of contention is the role epistemic probability plays in these arguments. You don’t need to be able to calculate precise probabilities in order to compare evidence between theism and naturalism and assess which is more likely. If your position is that epistemic probability doesn’t apply to theism, then you can do that. However, I think this would be a difficult position to defend insofar as without a nuanced understanding of philosophy of religion and probability, it could easily result in creating a straw-man version of both.
    .
    Aside from that, I can only reiterate my assessment that I think you’re depriving yourself of some very strong arguments against theism, so I hope you will keep them under consideration in the future.
    .
    Thanks again for taking the time to reply, I have enjoyed the exchange and it has even been a topic of conversion on Twitter which has not gone unappreciated.

  16. Monocle Smile says

    @EL

    Actually, if you listen to the past several weeks, it appears that you’re having an effect on Matt’s “supernatural” spiel. He’s even said that “supernatural” may be an incoherent concept since things that people say are supernatural would just be rolled into the natural world if we discover and learn about them. I’ll try to find an episode and time stamp, but I’m very sure of this.

    I guess the only valid definition of “supernatural” would arise if you define “nature” as encompassing only our local observable universe. However, I don’t see much point to using those definitions.

  17. Monocle Smile says

    @Chris

    Aside from that, I can only reiterate my assessment that I think you’re depriving yourself of some very strong arguments against theism

    We already have strong arguments, and the vast, vast majority of theists are bored shitless by this topic and/or don’t understand it at all. This is getting close to reverse-“sophisticated theology,” which is a topic I’d like to call in and talk about next week or so.

  18. says

    @EnlightenmentLiberal
    “I have found that Matt isn’t sufficiently Bayesian sometimes IMHO. (If you’re reading this Matt, I strongly suggest the book “Proving History” by Richard Carrier.)”

    I have a signed copy….which I’ve read.

    @Chris
    “You don’t need to be able to calculate precise probabilities in order to compare evidence between theism and naturalism and assess which is more likely.”

    I’ll refer you back to what I wrote: “implying that there’s some way to calculate or approximate which is more likely when one of the two proposed options (as far as I can tell) can’t be quantified or approximated”

    I’ve seen no demonstration of how to calculate OR approximate this.

    “If your position is that epistemic probability doesn’t apply to theism, then you can do that.”

    That seems to be the case.

    “Aside from that, I can only reiterate my assessment that I think you’re depriving yourself of some very strong arguments against theism, so I hope you will keep them under consideration in the future.”

    I’m aware that you think this, yet I don’t see that you’ve presented examples of the “very strong arguments against theism” that you claim I’m missing. As noted in my responses to the ones you did present…I not only didn’t find them strong, they barely seem to be a response to any typical fine-tuning argument.

  19. EnlightenmentLiberal says

    @Matt
    Let’s use your courtroom jury analogy. You say that a jury examines the claim “the defendant is guilty” and does not examine the claim “the defendant is innocent”.

    I agree that “not guilty” is not the same thing as “innocent”. I disagree when you say that the jury examines one question without examining the other question. A crucial insight of Bayesian reasoning is that you cannot examine a hypothesis in isolation. You are always comparing one hypothesis to another. A jury cannot decide that a defend is guilty without also examining the question “is the defendant innocent?”.

    To put numbers to it, one might say that you need to have 90% confidence that someone is guilty to render a guilty verdict. Thus anything less than 90% confidence is a “not guilty” verdict. However, from a “not guilty” verdict, I know that the jury believes that there is at least a 10% chance that the defendant is innocent.

    tl;dr
    You are wrong when you say that we can examine the question “is there a god?” without also examining the question “is it true that there are no gods?”.

    PS: I also agree that one is an atheist if one answers “no” to the following question “do you believe that there exists a god or gods?”.

    PPS: Thanks for your reply. I do look up to you. Keep up the great work.

  20. says

    @EnlightenmentLiberal

    “You are wrong when you say that we can examine the question “is there a god?” without also examining the question “is it true that there are no gods?”.”

    I’m sorry, but you’re just wrong in saying this…in part, because we’re NOT examining a question, we’re examining CLAIMS. The claim “some god exists” can be examined even if the claim “no god exists” cannot.

    The claim “this jar has an even number of gumballs” can be accepted or rejected, without ever considering the contrary proposition “this jar has an odd number of gumballs”.

    I can accept one or none of the possibilities. There may not be sufficient information to determine whether the number is even or odd – but the person who asserts that it IS even (or odd) has adopted a burden of proof. If you tell me the number is even, I can assess your evidence FOR that proposition, without ever being required to consider evidence for the contrary (which there may not be).

    I can accept your claim that the number is even, or reject it. Rejecting your claim does not mean that I’m convinced the number is odd, it doesn’t mean that I think it’s more likely that the number is odd, it doesn’t mean that I’m “leaning” toward the chance that it’s odd and it doesn’t mean that I’m 10%, 20%, 40%, 60% in favor of it being odd.

    It just means that you haven’t convinced me that it is rational to accept the proposition that it is even.

    Stop analyzing questions and start analyzing propositions and you’ll stop confusing yourself.

  21. frankgturner says

    Since Matt is reading this, I will just make a point that I often make (I should call the show about this). At several points you talk to the callers about what is “true.” Based on the context of what you were discussing, I think it would have made sense to discuss what is “factually correct.” I am among the only people I know who makes a distinction between “truth” and “factual correctness,” but I have been making that distinction a lot among believers lately (creationist protestants mostly, Catholics already get this they just dont use the terminology). Making the distinction can sow seeds of doubt.

  22. Conversion Tube says

    Listening to the call and what Chris has said here I think Chris is saying we can and sometimes should put on the theist hat and show them they are wrong even if you accept the first claim (that we can attempt to calculate the probability).

    That there is more than one way to skin a cat and even when we accept a few of their premises for the sake of argument, that their arguments STILL fail to be naturalism and it can sometimes get the point across better when we do it this way instead.

  23. frankgturner says

    @ 26
    Interesting thought CT. I will have to listen to it again with that in mind.

  24. Conversion Tube says

    Further to the discussion between Matt and EL.

    Funny I say this after my above post.

    “””PS: I also agree that one is an atheist if one answers “no” to the following question “do you believe that there exists a god or gods?”.

    As atheists We should be stating

    What is a “god”
    What are it’s attributes?
    How did you determine them?

    It’s our culture that leads us to forget to ask these first questions.

    It’s just assumed we have bought into the notion or idea of the god concept.

  25. frankgturner says

    @ Hippycow #7 and Matt D in general.
    1st of all, Matt you totally rock. I really wish I had known about you and TAE when I was much younger, but feces occurs. (Pardon my beef regarding the difference between what is “true” and what is “factually correct.” I have that with everyone and it comes from being raised among pseudo liberal Catholics).
    .
    IMHO WLC is not a “profesional philosopher” despite being labeled that way. He’s a pseudo politician looking for constituents. He is not interested in being “right” or “truth” in the factually correct sense. He is interested in his views being popular because it gives him power.

  26. favog says

    I thought Chris was saying that if you ask the theist if God has more options than random chance does, they’ll of course say God, and that’s where the idea that you can derive the larger number from, proceeding then to point out that it makes any particular universe less likely, including this one. Strikes me as more of a way to confuse the theist than to actually prove a point, but sometimes that’s worthwhile, too.

    Following arguments about the “supernatural” here and other places, I think that word belongs in the same garbage heap as “spiritual”. Nobody knows what it means, and trying to define it seems to just create pointless layers of noise that are way off whatever the original topic was. Surely everything that can be resolved by asking the questions “Is clairvoyance something supernatural” and “is there any reality to the supernatural” can be condensed to the much more efficient “Is clairvoyance something real?” and made a better question. Throwing away “supernatural” seems to me like the way to go.

  27. frankgturner says

    @favog #31
    I wholeheartedly agree that the term “supernatural” just creates noise and is a way of saying “I don’t know, therefore ______” (insert favorite explain without really explaining in the blank).
    .
    A lot of humans are insecure and don’t like not having explanations so they make things up to feel more secure. Some learn to get past that but a lot don’t.

  28. EnlightenmentLiberal says

    @Matt
    Thanks for the response. I think we’re close to figuring out our disagreement here, but I don’t think we’ll get much farther.

    The claim “this jar has an even number of gumballs” can be accepted or rejected, without ever considering the contrary proposition “this jar has an odd number of gumballs”.

    False.

    I can accept one or none of the possibilities.

    True.

    There may not be sufficient information to determine whether the number is even or odd

    True

    – but the person who asserts that it IS even (or odd) has adopted a burden of proof.

    True.

    If you tell me the number is even, I can assess your evidence FOR that proposition, without ever being required to consider evidence for the contrary (which there may not be).

    False. This is your critical error. Either you don’t understand Bayesian reasoning, or you reject it. I’m not sure which. A central tenant of Bayesian reasoning is that it is impossible to assess evidence for any proposition without also considering how that evidence might affect the likelihood of alternative hypotheses. Your estimation of the likelihood of the truth of a proposition always depends on the expected likelihood that we should have this evidence if the proposition is true, and on the expected likelihood that we should have this evidence if the proposition is false. That’s what Bayesian reasoning is. (Of course, you can always split the hypothesis “the proposition is false” into multiple sub-hypotheses. Bayes equation allows for this.)

    Do you agree that your assertions are contrary to Bayesian reasoning?

    If yes, do you then reject Bayesian reasoning as sometimes inapplicable or an incorrect way to do reasoning?

    If you think you are applying proper Bayesian reasoning, I strongly suggest that you re-read “Proving History”, or otherwise refresh yourself with basic Bayesian reasoning.

    I can accept your claim that the number is even, or reject it. Rejecting your claim does not mean that I’m convinced the number is odd, it doesn’t mean that I think it’s more likely that the number is odd, it doesn’t mean that I’m “leaning” toward the chance that it’s odd

    Correct.

    and it doesn’t mean that I’m 10%, 20%, 40%, 60% in favor of it being odd.

    False. In the language of Bayesian reasoning, 50% confidence aka 50% probability means “I don’t know”, 40% means “I lean slightly towards it being false”, 60% means “I lean slightly towards it being true”, 0% means “I am absolutely convinced that it is false”, 100% means “I am absolutely convinced that it is true”, etc.

    If you do not accept that the number of balls is odd, and let’s pretend the threshold for acceptance of a claim is 90%, that means that you have a 0% to 90% confidence, or Bayesian probability, that the number of balls is odd. From that, it necessarily follows that and you have a 10% to 100% confidence that the number of balls is even. (Assuming that “even” and “odd” are mutually exclusive and exhaustive.)

    It would be logically inconsistent to say “I do not accept that the number of balls is odd” and say “I am very convinced that the number of balls is not even”. By knowing something about your position about the proposition “the number of balls is odd” – by knowing that you do not accept the proposition “the number of balls is odd” – I know something about your position w.r.t. the proposition “the number of balls is even”. I know that you are not convinced that the proposition “the number of balls is even” is false, to exactly the same extant, the same confidence, that you are not convinced that the proposition “the number of balls is odd” is true.

    Again, in Bayesian reasoning, when you say something is true or false, and to what confidence, you are always comparing it to some alternative. If you have no alternative, then you cannot use Bayesian reasoning to say it’s true or false to any degree of confidence. And of course, most of the point of the book “Proving History” was to argue for the point that all proper empirical reasoning is Bayesian.

    IMHO, this plays nicely into my positivism that if you do not have some (observable) alternative, then the claim itself is not well formed and the truthness / falseness is not defined for the proposition. This is represented in Bayes equation thus: If you have no alternative hypothesis, then Bayes equation is inapplicable.

    @Conversion Tube

    PS: I also agree that one is an atheist if one answers “no” to the following question “do you believe that there exists a god or gods?”.

    As atheists We should be stating
    What is a “god”
    What are it’s attributes?
    How did you determine them?

    I think that the working definition of “god” is clear enough for me to say that there are probably no (observable) gods. I know some people think that I’m wrong – some people think that there is an observable god, and some people think that we don’t have a clear enough working definition of the concept of “god”. Obviously, I disagree.

    @favog

    Following arguments about the “supernatural” here and other places, I think that word belongs in the same garbage heap as “spiritual”. Nobody knows what it means, and trying to define it seems to just create pointless layers of noise that are way off whatever the original topic was. Surely everything that can be resolved by asking the questions “Is clairvoyance something supernatural” and “is there any reality to the supernatural” can be condensed to the much more efficient “Is clairvoyance something real?” and made a better question. Throwing away “supernatural” seems to me like the way to go.

    and

    @frankgturner in 33

    ~heart~
    (Too lazy to look up the escape sequence for less-than symbol.)

  29. Hippycow says

    @EnlightenmentLiberal #34
    I think the problem here is that Matt is addressing the claim that “it is true” and what he said about that is correct.

    Bayesian analysis doesn’t answer such questions. You can use Bayesian analysis to assign a probability for how viable a claim is based on background information, etc., but that is not the same as saying the claim is true or false. No matter how high or low the probability comes out, the question of whether the claim is actually true or false remains unresolved, until some actual concrete evidence to support it is available.

  30. ironchops says

    I need some help here on this Fine tuning argument thing.
    So far we (the people (animals) of planet earth) have found no concrete evidence of life anywhere except here (earth). Using math we have determined some sort of probability that life could occur out there somewhere but that’s all we have for now as far as I know. Question 1. How can the universe be fine tuned without life popping up all just any ole place. Question 2. Isn’t the universe constantly changing (tuning)? If so then Question 3. When will it become so fine tuned that we can’t live in it any more?

  31. EnlightenmentLiberal says

    @Hippycow
    Again, I have to politely but firmly disagree with everything you just said. I also strongly suggest that you read (or re-read) Richard Carrier’s book “Proving History”.

    This is how I would respond in short.

    You are wrongly creating a dichotomy between “demonstrated that a claim is true” and “assigning a probability about the truth of a claim”. This looks very much like a fallacious appeal to absolute certainty. In proper empirical reasoning, no claim is held to absolute certainty. In proper reasoning, every time you assert that some empirical claim is true, or you demonstrate some claim is true, no matter how much evidence you have, it is not beyond all doubt. Instead, you have merely raised our estimation of its truth to a very high probability or confidence, and that’s exactly what Bayesian reasoning is for.

    My other complaint is specific to ignoring alternative hypotheses. It’s hard to see on the “even vs odd gumball” example, so I’m going to use a different example. If you want to make some claim about the way the world is, and you cite some piece of evidence as supporting your claim, that alone is not sufficient to actually demonstrate that your claim is likely true, no matter how much evidence you obtain in this fashion. I don’t care how much evidence you get in favor of a proposition – at no point does that chain of reasoning end with the conclusion “and thus it is (likely) true”.

    Let’s take an example. Suppose you argue that a dog ate your homework. You supply a written statement of yourself as a witness that the dog ate your homework. You supply other signed witness statements. You supply a video of some dog eating some paper. You supply all sorts of other evidence. You argue that all of this is highly supportive of the proposition that a dog ate your homework. However, this reasoning on its own will never be sufficient to demonstrate that a dog actually ate your homework. To do that, you also need to address alternative hypotheses, and how much the evidence fits those other hypotheses. For example, I can imagine a simple conspiracy of you and your friends to concoct the written statements and video evidence. You then have to properly argue that the available evidence is more expected on your hypothesis compared to the alternative hypothesis(es). Until you do that, you have demonstrated nothing. That’s the key takeaway of proper Bayesian reasoning which both you and Matt currently do not understand.

  32. EnlightenmentLiberal says

    I didn’t phrase that quite right. I missed an important point.

    One of the central problems is this: You might argue that some piece of evidence strongly supports your proposition, and you might say you can do that without examining alternatives, but that is wrong. What you are actually doing is arguing that your piece of evidence is strongly expected if your proposition is true, but that’s not the same thing as arguing that your piece of evidence supports your proposition. The math of Bayes equation formally specifies this, but I’ll try it in English. In short, you might be right that your piece of evidence is strongly expected if your proposition is true, but that piece of evidence might be equally expected if some alternative proposition is true. If your evidence is equally expected if your proposition is true compared to if it is false, then your evidence does not favor your proposition.

    For example, I might claim that a dog ate my homework in front of my friends and video recorder. I might argue that my signed testimony of my friends, and my video evidence, is strongly expected on the hypothesis that my dog ate my homework. However, the professor would be right to argue that on the hypothesis that my friends and I are lying and conspiring together, the evidence is also quite strongly expected, and thus the evidence is not really good evidence for either proposition. To finally reach the conclusion that my dog ate my homework requires evidence that is strongly unexpected on alternative propositions, or enough evidence that is mildly more expected on my proposition, or some combination thereof. Again, all perfectly well modeled in the math of Bayes equation.

    For example, the professor might have secretly installed a listening device on my person, and this captured audible conversation of my friends and I engaging in conspiracy to manufacture testimony and evidence. This new piece of evidence, the audio recording, is strongly expected on one hypothesis and not the other, and for that reason this piece of evidence actually supports one of the hypotheses over the other.

    Again, the conclusion is that you can only determine that a piece of evidence favors a proposition by comparing it to alternative propositions.

  33. EnlightenmentLiberal says

    Ack. One more.

    The problem with the gumball example is that it’s hard to express this to someone who does not already understand.

    For the “dog at my homework” example, I gave several pieces of evidence which are roughly equally expected on the proposition and its alternatives. That means that the evidence is formally a non-sequitir if we restrict our analysis to just those two propositions (“the dog ate my homework” and “I’m lying that the dog ate my homework”), but it doesn’t feel like a non-sequitir. Our intuitions say that’s related.

    Whereas, for the gumball example, it’s much harder to come up with a piece of evidence which is equally expected on the even hypothesis and the odd hypothesis. Most kinds of evidence which are equally expected will seem much more like non-sequitirs. For example, if the even hypothesis is true, I expect that the Sun will rise tomorrow. This is a completely accurate prediction of my hypothesis and the background evidence, but it’s also a non-sequitir because it’s just as likely on the odd hypothesis. Our intuitions say that this is fallacious because it’s unrelated, but out intuitions say that the dog video evidence is not unrelated and not fallacious – but that’s just an example of our intuitions getting it wrong. Both pieces of evidence should not not affect our estimation of the truth of the proposition, and thus they’re both equally non-sequitirs, but it seems at some intuitive level that one is a non-sequitir and one is not. Resist that bad intuition.

    The only reason why some piece of evidence might favor the odd gumball hypothesis is because it’s expected on the odd hypothesis (ex: “I counted, and it is odd”), and it’s highly not-expected on the even hypothesis (ex: “I counted, and it is odd”). You need both prongs to properly conclude that it’s actually evidence in favor of your proposition, and not a non-sequitir like saying “I used my odd gumball hypothesis to predict that the Sun will rise tomorrow, and I observed that the Sun did indeed rise as expected”.

  34. EnlightenmentLiberal says

    PS: Technically, we could argue about the relative conditional probability, the expectedness, of the dog video evidence on both hypotheses. It’s not strictly a non-sequitir because it may be right to argue that it’s more expected on one hypothesis that the other, which means my example is bad. It’s hard to come up with examples on the fly. Sorry. Still, I think my point stands that we should be wary of evidence which seems related, but which should not change our estimation of the truth of the proposition, and to do that properly, we need to examine the evidence on the proposition and on the alternative propositions.

  35. frankgturner says

    @ EL
    Look I think I get where you are coming from in that if we exclusively use pure Bayesian mathematics to the exclusion of all of our other facilities as human beings such that we can state that we did so with pure Bayesian reasoning then you are correct that we cannot make a statement regarding the empirically factual correctness of said statement but only the probability of said hypothetical occurring. (You will have to excuse me in that I have not read Carrier’s book). Does it really matter though? What is the practical application of being technically correct and completely factually accurate here? (I know this sounds strange coming from someone who has a beef with the hosts not distinguishing “truth” from “factual correctness,” but even I see that sometimes it doesn’t matter).
    .
    IMHO you are beating a dead horse here. If it is important then Matt can address it. You are a good guy with a good mind and a lot of good ideas but I think you are trying to hard to be absolutely accurate.

  36. EnlightenmentLiberal says

    @frank
    I don’t think this is an esoteric point. I know I sometimes harp on esoteric points, but this is not one of them. This concerns the bedrock of what constitutes proper empirical and scientific reasoning. This is one of the most important things of all.

    When Matt uses his courtroom analogy and says that the jury can issue “guilty” without examining the question of innocence, that’s simply wrong.

    When you analyze a claim, any claim, you always have to look at all (plausible) alternative claims. A material fact does not count as evidence for a proposition unless it’s more expected or more likely on the hypothetical that proposition is true when compared to all alternative propositions. That’s simply how all proper empirical reasoning is. Matt often gets that wrong, and it’s irritating. It’s irritating because he is wrong, and it’s more irritating because he is teaching some people to think badly.

    One conception of “the burden of proof” is just a restatement of some of the things I’ve just said. Other conceptions of the burden of proof are a cultural contrivance to put the onus on producing evidence on the person making the claim in order to aid fairness and to help avoid the situation where some asshole wastes another person’s time. When you analyze a claim to see if it’s true or false or “I don’t know”, and to what degree of confidence, it doesn’t matter who provided the evidence. Rather, the concept “the burden of proof” is AFAIK often used merely as a weapon against someone who wants you to do their work for them in the argument, and as a weapon against someone wasting your time when they have an bullshit unsupported claim. This conception is merely a heuristic that saves time by sacrificing some degree of knowledge. It’s to avoid analysis-paralysis.

  37. John Iacoletti says

    I’m not following you here at all. If you’re a prosecutor and you claim the defendant is guilty, and then fail in your burden of proof to show that he is guilty, you have done nothing whatsoever to demonstrate that he is innocent with any probability or confidence.

  38. frankgturner says

    And I get that people do in their real lives consider the possibility of innocence when analyzing whether people are guilty or not, even without realizing that they are at times. The concepts of actus reus and mans rea exists due to critical examination of innocence. One step at a time though. People have to be opened to thinking about other propositions even being possible before they can think about considering the infinite range of possibilities.
    .
    You are making a claim that he is teaching people to think badly. Really? What evidence to you have that supports this claim? Have you considered all possible ways in which people could be thinking?
    .
    I get that you are concerned about the ways in which people might be thinking and may even be convinced that they are thinking that way, but have you asked?
    .
    I don’t know if he is teaching people to think badly. He might be, but he might not. You seem awfully certain that he is for someone who claims an openness to other alternatives.

  39. Chris says

    There seem to be many hidden assumptions being made about the specifics of what theism entails or permits.

    Any response that attempts to argue that naturalism is actually more probable must buy into their hidden assumption that the probability of theism is somehow known.

    …appealing to incalculable probabilities isn’t really getting anyone anywhere.

    It’s magic. It could be argued that it predicts either nothing or nearly anything.

    I know Matt is partial to a quote by David Hume: “A wise man, therefore, proportions his belief to the evidence.” Well the question is, how do we do that? From his comments, I’m left wondering how exactly Matt determines whether or not a given hypothesis best explains the available evidence. I’m concerned that Matt’s dismissive attitude towards theism could easily be interpreted as attacking a straw-man, if not begging the question for naturalism outright.
    .
    When I appeal to epistemic probability, i’m appealing to a framework that carefully outlines the relationship between the evidence and a hypothesis, with a foundation in rigorous mathematical analysis. If a clear and consistent definition of evidence cannot be provided, then asking a theist to provide evidence for their claims becomes a fruitless endeavor.
    This outline should provide a better understanding of what i’m talking about: Link
    .
    Ultimately, my point is this: In the face of theistic claims, I think atheists should strive to offer more than an incredulous stare.

  40. EnlightenmentLiberal says

    @John

    I’m not following you here at all. If you’re a prosecutor and you claim the defendant is guilty, and then fail in your burden of proof to show that he is guilty, you have done nothing whatsoever to demonstrate that he is innocent with any probability or confidence.

    I don’t know if you’re a Bayesian, e.g. if you accept Bayesian reasoning as the only proper kind of empirical reasoning (give or take other forms of reasoning which are equivalent to or a subset of Bayesian reasoning). I’m operating from that assumption. Let me try to persuade you of this foundational fact.

    Naively, one might enter a jury with 50 50 odds that the person is innocent. A property of Bayesian reasoning is that we always have prior estimates of the likelihood of a hypothesis. Usually this likelihood is based on prior experience and evidence, which is itself based on previous applications of Bayesian reasoning. Of course, you have to start somewhere (IMHO), and generally the safe starting priors are 50 50 odds, e.g. 50% confidence, e.g. “I don’t know”.

    In the real world, I might be able to gather together evidence to inform my priors about the probability that any one person in front of a jury trial is innocent or guilty. I might be able to break it down by kind of crime – property crime, violent battery, etc. I might be able to break it down by racial lines. And so on.

    Let’s assume for the sake of argument that I know nothing about the probability distribution of guilt vs innocence for people that appear before a jury. As I said above, the prudent thing to do is adopt 50 50 odds that they are guilty or innocent, e.g. 50% confidence, e.g. “I don’t know”.

    Now, if the prosecutor fails to present any relevant or meaningful evidence, then my adjusted estimation of the likelihood of guilt remains unchanged from my priors. My final estimation is still 50 50 odds, e.g. 50% confidence, e.g. “I don’t know”. That’s just a straightforward and inescapable consequence of the math.

    However, suppose the prosecutor presents a piece of evidence X, and asserts that this piece of evidence means that the defendant is guilty (to a high degree of confidence). I’m not going to agree just based on a bald assertion. Instead, I need to examine whether the hypothesis that the defendant is guilty is concordant with the piece of evidence X. However, that alone is not enough. I also need to examine various hypotheses of innocent to see if this piece of evidence is concordant with innocence. If the piece of evidence is equally expected on some theory of innocence as it is on some theory of guilty, then the evidence does not adjust my final estimation of the likelihood of guilt one way or the other.

    For example, murder trial of a guy who purportedly killed his wife. Wife was killed by a knife. It was a kitchen knife from the house. The style matches the other knives in the silverware drawer of the house, and the serial numbers match meaning it’s from the same set as the other knives in the house. The prosecutor introduces a piece of evidence X: they found the husband’s finger prints on the murder weapon – not in blood, but just normal finger prints. The prosecutor claims this evidence supports the conclusion that the husband is guilty of murder. However, before the jury can make that assessment, they need to examine various theories of innocence and see if they can make the evidence fit a theory of innocence. In this case, the jury would realize that there are other reasons why the finger print might be on a knife from the household which have nothing to do with murder. In this case, the evidence is strongly expected on one particular hypothesis of guilty (if someone is stabbed, you expected that any finger prints on the weapon should match the killer), but the evidence is also strongly expected on one particular hypothesis of innocence (you also should expect that kitchenware in the house will commonly have finger prints of members of the household, and maybe the killer wore gloves). Thus, this finger print evidence is very bad evidence for guilt, or not evidence at all. That is the proper conclusion, and the only way to reach that conclusion is to examine all (plausible) alternative hypotheses.

    You cannot establish any claim in a vacuum. You are always establishing probabilities of empirical truths compared to alternative hypotheses. This is a deep, insightful, and very important realization. It’s also a baby step away from many of the important realizations of positivism: an empirical claim is meaningless if there is no alternative hypothesis you can test it against, and all empirical claims are meaningfully true only to the extent that an empirical claim can be differentiated by testing from alternative testable claims.

    @frank

    You are making a claim that he is teaching people to think badly. Really? What evidence to you have that supports this claim? Have you considered all possible ways in which people could be thinking?
    .
    I get that you are concerned about the ways in which people might be thinking and may even be convinced that they are thinking that way, but have you asked?
    .
    I don’t know if he is teaching people to think badly. He might be, but he might not. You seem awfully certain that he is for someone who claims an openness to other alternatives.

    I don’t know what that last sentence means. It seems you’re doing equivocation, or you are confused. I am arguing that we need to examine all (plausible) alternative propositions before we can properly conclude that a piece of evidence actually supports a proposition. That’s fundamentally different than being open to alterantive ways of thinking and reasoning. I fail to see a connection except for a shared word.

    I also think it rather obvious that if Matt says that you can determine guilt without examining the question of innocence, then Matt is teaching people to think badly. The straightforward understanding of that claim is simply and completely wrong, and I think it rather reasonable to understand that some people will be swayed to agree with the straightforward but wrong understanding of the claim.

  41. John Iacoletti says

    I accept the methodology but disagree on what it is that’s going on. In the murder case, I would claim that you’re not evaluating an alternative hypothesis of innocence, you’re instead evaluating whether the presented evidence is sufficient to warrant the conclusion that the person is guilty — is “he murdered her” the only (or even best) explanation for fingerprints on the knife? That’s not an argument in favor of innocence. He could have gotten the fingerprints on the knife from living there and handling the knife at dinner, and still killed her with that knife. Or killed her with a different knife that he wiped clean. Failing to prove a person is guilty doesn’t affect your probability or confidence that the person is innocent in any way.

  42. EnlightenmentLiberal says

    @John

    Failing to prove a person is guilty doesn’t affect your probability or confidence that the person is innocent in any way.

    That’s just logically inconsistent.

    Let me try to use the language of rational gambler. As a rational gambler, I would have a certain estimation that the person was guilty. This estimation, or odds, that the person was guilty is necessarily the exact compliment of the odds that the person is innocent. I know that the probability that the person is guilty plus the probability that the person is innocent equals 1, e.g. 100%.

    P(guilty) + P(innocent) = 1.

    That’s simply part of what it means to be rational and to not hold conflicting beliefs. You cannot have 50-50 odds that someone is innocent and also have 60-40 odds that the same person is guilty. That’s logically inconsistent and irrational. If you are rational, your belief concerning the odds of guilt necessarily entails a belief concerning the odds of innocence.

    In English, it’s illogical and irrational to say any of the following:
    “I’m not yet convinced that the person is guilty, but I believe the person is innocent.”
    “I’m not yet convinced that the person is guilty, but I believe the person is not innocent.”

    That necessarily means that if you are not convinced that the person is guilty – if you do not know if the person is guilty, then it necessarily follows that you do not know if the person is innocent. One belief constrains the other.

    If I currently do not accept that the person is guilty, that means that my belief concerning the odds that the person is guilty is constrained in some way. For example, it might mean that I wouldn’t take a bet with a certain odds over a certain threshold. If I am a rational gambler, that necessarily constraints my beliefs concerning which bets I should take for the bet that the person is innocent.

    I know this is how the show has addressed the topic for time immemorial, but the conventional wisdom of the show is dead wrong.

    I’m mostly just repeating myself now. I hope the gambler language helps. Otherwise, I don’t know how to respond to something so obviously and blatantly wrong.

  43. EnlightenmentLiberal says

    Correction:

    “I don’t know if the person is guilty, but I believe the person is innocent.”
    “I don’t know if the person is guilty, but I believe the person is not innocent.”

    “I am not yet convinced that X is true” acts as a way to advertise that you believe the odds are constrained on one side. If you need 1 to 9 odds to be convinced, that corresponds to a probability of 90%, (1 / (1 + 9)), and thus saying “I am not yet convinced that X is true” tells me that whatever odds you give to the event are in the range of 0% to 90%. And we all give odds to events. It’s part of being a rational agent who does risk-benefit analysis. You might not realize that’s what you’re doing, but you’re doing it, perhaps informally and perhaps partially unconsciously.

  44. Monocle Smile says

    @EL

    I also think it rather obvious that if Matt says that you can determine guilt without examining the question of innocence, then Matt is teaching people to think badly. The straightforward understanding of that claim is simply and completely wrong, and I think it rather reasonable to understand that some people will be swayed to agree with the straightforward but wrong understanding of the claim

    I think this is an extremely unfair extension of an analogy that was never meant to be taken this far.

  45. EnlightenmentLiberal says

    @MS
    How do you explain the following written by John from the show?

    Failing to prove a person is guilty doesn’t affect your probability or confidence that the person is innocent in any way.

    I don’t think I’m taking this out of context, or extrapolating badly, or making a bad analogy, etc. I’m taking this at face value.

  46. Monocle Smile says

    I don’t agree with that statement from John, but I find your bringing that up to be a bit of a non sequitur. Not sure what that has to do with what you said before.

  47. EnlightenmentLiberal says

    @MS
    I see a consistent pattern from several hosts of the show, especially Matt, which displays a strong lack of understanding of proper empirical reasoning, e.g. Bayesian reasoning. I used that quote from John as an example of the trend. For further evidence that I am not misrepresenting Matt, John, and others, could you please re-read post 24 by Matt in this thread. I think Matt makes this (wrong) claim very clearly. Especially this bit:
    Context: Matt is talking about whether the number of gumballs in a jar is even or odd.

    I can accept your claim that the number is even, or reject it. Rejecting your claim does not mean that I’m convinced the number is odd, it doesn’t mean that I think it’s more likely that the number is odd, it doesn’t mean that I’m “leaning” toward the chance that it’s odd and it doesn’t mean that I’m 10%, 20%, 40%, 60% in favor of it being odd.

    As I tried to explain to John just now, that’s just logically inconsistent and irrational. Of course rejecting the claim that the number is odd necessarily means that you believe the odds that the gumball number is odd is 50-50 or worse, and to be logically consistent this entails that you must believe that the odds that the gumball number is even is 50-50 or better. That’s what it means to be rational.

    Again, I think that part of this might be a misunderstanding of the math. Again, 50% Bayesian confidence in a proposition X is not “I lean towards X being true”. In Bayesian terminology, a 50% confidence level is a 50% probability, which is the same thing as 50-50 odds, which is the same thing as lacking an affirmation or positive belief that it’s true and lacking an affirmation or positive belief that it’s false.

  48. EnlightenmentLiberal says

    Ack, that is ambiguous.

    If someone rejects a claim X, assuming honesty, I can ascertain that this person has a certain belief concerning the odds of the truth of the claim X, and that these odds lie between 0% and 50%, e.g. between 1 to infinity odds against and 1 to 1 odds. It may be that they are undecided about the proposition, which is expressed as 50% probability or 1 to 1 odds. If may also be that they lean against the truth of claim X, which would put their confidence at the appropriate spot between 0% and 50%, e.g. between 1 to infinity odds against and 1 to 1 odds. I don’t mean to say that the person necessarily believes the truth of the matter may be anywhere between 0% and 50%. Rather, that statement merely allows me to constrain the set of possible positions which the person can hold.

    Regardless, if the person is rational, and if they have looked into the claim X with any detail, it means that they have formed a belief concerning the probability that it’s true. It may be as simple as “I don’t know”, which is 50% probability, e.g. 1 to 1 odds. They may believe it’s true to a high degree of confidence, i.e. 90% probability, e.g. 9 to 1 odds for.

    Also regardless, if the person is rational, their confidence, their estimation of the probability, their estimation of the odds that X is true must match their confidence, their estimation of the probability, their estimation of the odds that X is false.
    P(X) + P(not X) = 1
    That’s an indispensable part of what it means to be rational. To be a rational agent, you need to estimate odds of almost everything (again perhaps informally and on demand), and a perfectly rational agent would have no logical contradiction between their beliefs, such as a belief that it’s 50-50 odds that X is true and 40-60 odds that X is false. That’s what we should aspire to as part of Matt’s slogan “I want to believe as many true things as possible and as few false things as possible”.

  49. Bugmaster says

    @Matt:

    I have a question about your take on methodological naturalism; sorry if this is a little off-topic, but I think the matter is at least tangentially related to the discussion here.

    You often say that methodological naturalism merely rules out our ability to study the supernatural. It could still exist, but we can’t understand it scientifically until someone can provide some evidence for its existence. That’s fine as far as that goes, but, as far as I understand, we study things by looking at how they interact with other things.

    For example, we can study the motion of a thrown rock by observing it “directly”; but what that really means is that photons are coming from the Sun (or some other light source), hit the rock, reflect off of it, and are absorbed by our retinas (unless, of course, you happen to be standing directly underneath the rock when it falls, in which case a whole other kind of interaction takes place). We cannot observe magnetic fields in this way, but we can observe their effects on e.g. metal filings. We cannot observe love (or hatred), but we can examine the behaviour of people who are under the influence of this emotion, and infer its presence — just as we did with the invisible magnetic field.

    So, when you are saying, “the supernatural cannot be studied, even in principle”, are you not really saying, “the supernatural has no effect on absolutely anything” ? How is that different from saying, “the supernatural does not exist” ?

  50. frankgturner says

    Ok EL, it seems you are not understanding where I am coming from nor where Matt is coming from. Let me try to pin it down.
    .
    A Lot of theists and presuppositionalists (StB comes to mind) make unfounded claims without ANY evidence to support their claim. Like WLC they act very confident of themselves and may give an air of confidence that their constituents seem to enjoy (as I state, I use the term constituents as I think WLC is more of a politician).
    .
    The factual correctness of their claims or any mathematical Bayesian reasoning may not matter to the constituents who are often driven by emotional thinking, not logical or rational thought. A first step in to how to think well seems to be about pulling them AWAY from that. For those on the fence, using SOME logic and appeal to reason makes sense. It may not have to be fine tuned to perfectly accurate logic in doing so. I get what you are talking about in a mathematical sense and the logic and the numbers do make sense to me and are rational. Human beings are unfortunately NOT purely mathematical beings that do things based on the logic of the numbers.
    .
    Theists know this and realize that gullible people can be driven to believe things that are not factually correct by sounding confident about them. Pardon me for Godwin here, but Hitler knew this, which is why he talked about a lie that was big enough and repeated often enough with enough confidence being used to confidence people that the lie was right. Courtroom attorneys know this too. They often don’t pick (or want to pick) people who are well educated to serve on trials. They want people who are swayed by emotion, not logic.
    .
    I know that a person who is declared “not guilty” is not necessarily “innocent.” Those are separate principles (though if we are going to discuss Bayesian statistics, this is a situation where there probability of “not X” can, in some instances, influence the probability of ‘Y,’ but that is separate from the point that I am trying to make).
    .
    As Matt has said in previous shows, the point of the courtroom analogy is to demonstrate that the person who has the burden of proof is the one making the positive claim, which is the accuser, the plaintiff. A defending attorney does not need to proof that their client “is innocent,” only to debunk the claim that the client “is guilty.” The default position is that they are “not guilty” and that i reflected in the verdict. No one is every declared “is innocent” in a verdict. If you can’t provide evidence for a God existing the default position is that he is “not guilty” of existing. Matt goes further to talk about his belief that God is innocent of existing but that is treated separately.
    .
    Now to make a connection to what I was saying before, From my experience listening to Matt I think it is that case that theists don’t want you to be opened to examining all plausible alternatives. Theists know that they have no solid empirical evidence upon which to back their claims, so they want you to accept something that is not really evidence to support a claim as though it were evidence to support a claim and close your mind to thinking otherwise. Before we examine all (plausible) alternative propositions before we can properly conclude that a piece of evidence actually supports a proposition you have to get people to examine alternative propositions. THAT is what I think Matt is trying to do.
    .
    Once they are opened to that THEN maybe they will accept that what they perceive as evidence of a claim is not really evidence of that claim. Sort of like the watch on the beach analogy (creationists fail to realize that the beach is FILLED with watches).
    .
    One of the things that I see many believers failing to understand is that their scripture can have value as a learning and teaching tool and have some great insight onto the human condition WITHOUT it necessarily equating with factual correctness. Many a believer I have met fails to realize that there are moral lessons inside of Genesis that have just as much meaning, maybe even more so, if the stories are fictional parables not meant to be taken as fact. So they are “true” in a moral or philosophical sense without being “factually correct.” That is why I say that I want to believe as many “factually correct” things as possible and as few “factually incorrect” things as possible. In general I also want to believe as many morally and philosophically “true” things as possible and as few morally and philosophically “false” things as possible as well. However, sometimes something can be morally and philosophically “true” while at the same time being “factually INcorrect,” such as a parable.
    .
    If something can be factually incorrect, but true in a moral and philosophical sense, then can’t it be factually correct but morally false? I have come to think that what you are stating at EL is “factually correct,” but misses a key intellectual point in that Matt may not have been intended the way that you are stating his views. You are insisting that he is wrong and maybe in a technical sense you are right, but did he mean to extend the analogy to the point where you are taking it? Is that really putting what he is saying in the right context?
    .
    To get people thinking about the Bayesian logic and how to apply it (which you have correct) you FIRST need to get them thinking of alternative ways of thinking. There has to be some theistic deprogramming. Not everyone will automatically jump to this mathematical conclusion regarding Bayesian logic that you bring up based upon the courtroom analogy or the gum ball analogy. Yes if you want them to be technically accurate in a purely Bayesian sense they will have to understand the equations and logic behind it. They need to be OPENED to that idea FIRST. Having people go from theistic deprogramming to Matt Dillahunty programming doesn’t opened their mind to application of Bayesian reasoning. That is just trading one deity for another (pardon my using you as a deity in this analogy Matt to make the point).
    .
    If you get them opened to other ways of thinking first they can be opened to the ideas of the hosts being wrong about their applications of the ideas in certain scenarios. That is what I was getting at.

  51. Bugmaster says

    @EnlightenmentLiberal:

    I think the problem with applying Bayesian reasoning to theological claims is our lack of a common prior. An Bayesian atheist’s prior probability for a god’s existence is somewhere close to epsilon; but a believer’s prior is somewhere closer to (1-epsilon). Thus, they tend to interpret evidence in very different ways. You might argue that both people should start with the prior of .5 and then work things out from there, but, in practice, I don’t think this is humanly possible.

    That said, I think that Matt’s reasoning is correct with regards to the gumballs. He rejects the statement “the number is odd”, but this entire point is that this in no way implies the acceptance of the statement, “the number is even”. All he’s saying is, “I don’t know how many gumballs there are, and neither do you”.

    On the other hand, you are absolutely correct in saying that, under Bayesian reasoning, “I don’t know” translates into “P = 0.5”. I think this is, in fact, an important point; because it allows us to have different values for P besides 0 (“the count is even for sure”), 1 (“the count is odd for sure”), and 0.5 (“I don’t know”). Since beliefs lead to other beliefs, and since in real life we can’t be certain about anything, we can only reason reliably by using Bayesian belief networks. Syllogisms such as “If X then Y; X; therefore Y” are just approximations that mean something roughly like, “P(Y|X) >= (1-epsilon); P(X) >= (1-epsilon)”.

  52. frankgturner says

    @Bugmaster # 57
    Thank you, you made my point better than I did. Matt is not trying to get people to think in a way that is mathematically unsound or illogical or fails to recognize Bayesian equations or claim absolute certainty about a piece of logic that does not agree with a Bayesian equation. He is just trying to get people to open their minds. Opening your mind to the possibility that the courtroom analogy is not a good one because of a potential to incorrectly interpret a Bayesian equation is fine, that at least gets a person thinking. Not everyone comes to that incorrect conclusion (most people are not statisticians).
    .
    I have a response to your post in # 55. In the past primitive humans believed lightening to be supernatural, bolts thrown by the Gods in ways in which they could not understand. Due to scientific studies we came to comprehend differences in atomic and molecular charges, Brownian motion, etc. (http://en.wikipedia.org/wiki/Lightning). We came to understand how lightning works and recognize it as a natural phenomenon. Did that mean that lightning was supernatural until we understood it? Not really, it always was natural. You could say that it was supernatural, because we did not understand it. However, this would imply that once we understood it, it became natural.
    .
    That is not really true though, it always was natural, as humans we just did not understand it. We do not understand or can equate or measure (completely) love or hate (though we can measure brain waves and see patterns in those experiencing love and hate). Does not make them unnatural?
    .
    As Matt says words have meaning by consensus, so the meaning of the word “supernatural” may change over time. If you thin about it though, based on scientific principle, nothing supernatural really exists because just because something is beyond our understanding does not mean that it forever will be. And once we gain an understanding of it we may realize that it never was supernatural at all. So the word, “supernatural” for some means, “I don’t understand it right now even though I can observe it happening.” For others, it might mean, “I don’t understand it right now and don’t believe that I ever will be able to understand it.” For others it might mean, “I don’t understand it and don’t think I will be able to within this limited existence” (presuming that there is a demonstrable existence beyond this one).
    .
    It means a lot of different things to a lot of different people, but seems to hinge on the idea of the explanation being un-demonstrable or un-observable. Because science hinges on a dogma of the observable and demonstrable, supernatural claims are outside of science. For science to examine a supernatural claim it would have to be observable and demonstrable while still being considered supernatural (which sounds oxymoronic).

  53. EnlightenmentLiberal says

    @frank in 56
    Ok. I’m not sure if I have a response to a lot of that. A lot of that is good. I know that I’m sometimes bad at explanation, and I’m definitely bad at persuasion.

    @Frank, post 58

    He is just trying to get people to open their minds.

    I would prefer he not use bad epistemology to do so. One of the reasons this is a big sticking point for me is the whole supernatural discussion. Matt – at least before 3 episodes ago – seems perfectly willing to say that evidence can support the assertion that something is natural, but is adamant in his assertion that evidence can never show something is supernatural. That right there is a reasoning fail.
    Further, I don’t think Matt is as productive as he can be, because I have seen many theists who feel that an atheist who adopts Matt’s particular reasoning to be dishonest, or at least wrong. Matt is wrong. The theist sees Matt demean the theist’s position because they have no evidence, then ask for evidence, and then not more than a few seconds later also say that no amount of evidence could ever show that the theist is right. That’s the clear unequivocable end result of what Matt is saying, and I want him to knock it off. It makes the rest of us atheists look bad.

    As for the rest of the post, you’re having a different conversation than Matt and I (mostly me). Matt has often claimed in no uncertain terms that no amount of evidence could ever possibly show that something supernatural exists. He has even given examples at times, such as Thor the thunderbolter. That’s simply wrong. You are having a different conversation. I think I agree in whole with your position w.r.t the supernatural.

    @Bugmaster

    I think the problem with applying Bayesian reasoning to theological claims is our lack of a common prior. An Bayesian atheist’s prior probability for a god’s existence is somewhere close to epsilon; but a believer’s prior is somewhere closer to (1-epsilon). Thus, they tend to interpret evidence in very different ways. You might argue that both people should start with the prior of .5 and then work things out from there, but, in practice, I don’t think this is humanly possible.

    If you think that’s not possible at all, even a little bit, then I suggest that you should work on your critical thinking skills, because it’s totally possible. In essence, you just asserted that atheists are just as dogmatic as theists, and that’s simply wrong.

    Quoting Bugmaster

    That said, I think that Matt’s reasoning is correct with regards to the gumballs. He rejects the statement “the number is odd”, but this entire point is that this in no way implies the acceptance of the statement, “the number is even”. All he’s saying is, “I don’t know how many gumballs there are, and neither do you”.

    Thus, when Matt makes the assertion “I do not accept that the number of gumballs is odd”, I can learn something about positions beliefs w.rt.. the claim “the number of gumballs is even”. This is contrary to what Matt has clearly stated. Matt is wrong. Matt could not have been clearer when he said it here:
    Bolding added

    I can accept one or none of the possibilities. There may not be sufficient information to determine whether the number is even or odd – but the person who asserts that it IS even (or odd) has adopted a burden of proof. If you tell me the number is even, I can assess your evidence FOR that proposition, without ever being required to consider evidence for the contrary (which there may not be).

    I can accept your claim that the number is even, or reject it. Rejecting your claim does not mean that I’m convinced the number is odd, it doesn’t mean that I think it’s more likely that the number is odd, it doesn’t mean that I’m “leaning” toward the chance that it’s odd and it doesn’t mean that I’m 10%, 20%, 40%, 60% in favor of it being odd.

    Most of that quote is defensible, except for the last part. Matt shoots himself in the foot when he gives that spread of percentages, including above 50% and below 50%. I don’t know what to make of that. Of course it’s inconsistent to say “I do not accept your claim that the number is even” and say “I believe that it is false that the number is odd”. By your own reasoning, if you say “I do not accept the claim that the number is even”, I do know something about your position w.r.t. the proposition “the number is odd”.

    I strongly suspect that Matt doesn’t understanding that 10% confidence in a proposition like “the number of gumballs is odd” means “I strongly believe that the number of gumballs is even”. The seemingly only alternative is that Matt would be ok with someone who 1- reject the “even” hypothesis and who 2- asserts that the “odd” hypothesis false. I hope everyone can see that this combination is obviously ridiculous. Or something.

    PS:
    To put it another way, it is improper reasoning to accept something as true if you only hear one side of the argument. You have to hear both sides before before you are justified to make a decision. For example, it would be highly improper to merely consider the prosecutor’s argument that the fingerprint on the murder weapon is consistent with a theory of guilt without also considering other explanations of how that finger print might have gotten there, and the relative plausibility or probabilty of the competing theories.

    To put it another way, something is not evidence for a proposition unless it’s more likely on that hypothesis than on all alternatives.

    Bayes equation formally captures all of this.

  54. Bugmaster says

    @EnlightenmentLiberal #60:

    In essence, you just asserted that atheists are just as dogmatic as theists, and that’s simply wrong.

    Well, firstly, I think that most people are fairly dogmatic about stuff like this, atheists are no exception — though, obviously, religious converts and de-converts do exist, so it is not the case that everyone is dogmatic. But secondly, given the sheer complexity of the world, I don’t think that we mere mortals simply have enough time to work out every belief from first principles; religious beliefs are no exception.

    Thus, when Matt makes the assertion “I do not accept that the number of gumballs is odd”, I can learn something about positions beliefs w.r.t. the claim “the number of gumballs is even”.

    Right, the gumball scenario is a special case, because we know a priori that there are only two possibilities: either the count is even, or it is odd. However, consider a different scenario. Let’s say that I propose a bet, as follows: “I claim that the remainder of dividing the count by 10 is 1. As you and I are both aware, neither of us has any special knowledge about the count; still, I just feel deep in my soul that the remainder is 1. If I am right, you will pay me $X. If I am wrong, I pay you $Y”. Will you take the bet if X = Y = 1 ? Are there any values of X and Y for which you would (or would not) take the bet ?

    I strongly suspect that Matt doesn’t understanding that 10% confidence in a proposition like “the number of gumballs is odd” means “I strongly believe that the number of gumballs is even”.

    Yeah, but I don’t think this is what Matt meant to say. I think you’re kinda splitting hairs here. That said, I’m not Matt, so I don’t feel right speaking for him…

    You have to hear both sides before before you are justified to make a decision.

    In most cases in the real world, there aren’t just two sides to the story; there are, in fact, an infinite number of sides. That is why the Bayes formula has a term like “P(A|~B)” in it, where “~B” stands for “all of these other alternative theories”. To use your analogy, if we can estimate that the probability of a fingerprint being on that weapon for reasons other than the obvious (i.e., the defendant touched the weapon) is very small, then we don’t need to enumerate all of the possible ways the fingerprint could’ve gotten there (alien pranksters, divine intervention, CIA agents with fingerprint faking kits, software glitch, mass hallucination, …). All we need to do is say something like, “P(fingerprint | ~touch) <= 0.01", and leave it at that (obviously I made up the number 0.01, but you get my point).

  55. says

    @Chris
    “Ultimately, my point is this: In the face of theistic claims, I think atheists should strive to offer more than an incredulous stare.”

    So, instead of actually addressing the objections…instead of actually presenting the better arguments that I’m missing, you imply that what I’m doing is comparable to an incredulous stare?

    Done here.

  56. EnlightenmentLiberal says

    @Bugmaster
    You’ll have to forgive me. I have a higher estimation of the atheist population. I think by far atheists tend to be less dogmatic than religious people concerning their position w.r.t. gods. Of course, there are still plenty of atheists who are dogmatic about some issues, gods included.

    The gumball scenario is not a special case. We live our life according to alternative hypotheses. Oftentimes we omit implausible or obscure scenarios, but that’s a heuristic which we all develop to avoid analysis-paralysis. I’m not demanding full formal Bayesian analysis. I’m just demanding that we learn it, understand what constitutes good and bad reasoning, and try to approximate it in our daily life, subject to all of the usual heuristics.

    I claim that the remainder of dividing the count by 10 is 1. As you and I are both aware, neither of us has any special knowledge about the count; still, I just feel deep in my soul that the remainder is 1. If I am right, you will pay me $X. If I am wrong, I pay you $Y”. Will you take the bet if X = Y = 1 ? Are there any values of X and Y for which you would (or would not) take the bet ?

    Of course. The math is trivial. It’s 90% odds that the remainder is 1 (from some intended implicit assumptions of the scenario). So, I only need a bet with payout odds better than 1 to 9, or:
    payout / cost >= 9
    Example: If the bet is you pay me $9 and I pay you $1, then we should come out even over the long term. Anything with better odds will make me money.

    In most cases in the real world, there aren’t just two sides to the story; there are, in fact, an infinite number of sides. That is why the Bayes formula has a term like “P(A|~B)” in it, where “~B” stands for “all of these other alternative theories”. To use your analogy, if we can estimate that the probability of a fingerprint being on that weapon for reasons other than the obvious (i.e., the defendant touched the weapon) is very small, then we don’t need to enumerate all of the possible ways the fingerprint could’ve gotten there (alien pranksters, divine intervention, CIA agents with fingerprint faking kits, software glitch, mass hallucination, …). All we need to do is say something like, “P(fingerprint | ~touch) <= 0.01", and leave it at that (obviously I made up the number 0.01, but you get my point).

    Fully agreed.

  57. Bugmaster says

    @EnlightenmentLiberal #63:

    Regarding the dogmatism of people, I am not convinced. To be precise, I wouldn’t necessarily say that most people are “dogmatic”, exactly; it’s just that most of us have mental biases (one way or the other) that are built into us by billions of years of evolution, and centuries of culture; in Bayesian terms, we have built-in priors. It is unreasonable, IMO, to say, “just ignore your biases”, because human minds do not work that way.

    Regarding my modified gumball scenario, as well as my paragraph you quoted at the end: my point there was that the original gumball scenario is a bit artificial. If someone convinces us that P(even) = 0.1, we can automatically assume that P(odd) = 0.9, because those are the only two choices. But that kind of reasoning does not work if there are more than two choices (i.e., you cannot privilege any of the remaining choices); and, in the real world, there are often many more choices than that. But it sounds like you agree, so I guess this entire discussion was superfluous 🙂

  58. EnlightenmentLiberal says

    @Bugmaster
    I agree people are biased. Saying that someone has a prior of epsilon, e.g. a very very small number, or 1 – epsilon, a number very very close to 1, is different than saying people have biases. Of course people have biases, but biases can be overcome with additional evidence and with training to recognize biases. Whereas, it’s basically impossible to ever overcome a prior of zero e.g. epsilon.

    I agree that for most situations in the real world, there are more than two competing hypotheses. I also think it rather obvious that you can still apply the principles of Bayesian reasoning in those cases. A piece of evidence counts as evidence in support of a hypothesis only if there is no competing plausible hypothesis which fits the evidence equally well. Of course, words like “plausible” and “fit” are modeled mathematically in Bayes equation.

    Also of course, Bayes equation is subject to garbage-in garbage-out. However, IMHO it’s still useful. Just like the premises of a formal syllogistic argument are unsupported, it can be useful to construct such arguments in order to pinpoint the points of disagreement, making a Bayesian argument can be useful to pinpoint the areas of disagreement.

  59. frankgturner says

    @ EL
    For starters, regarding #58, I WAS having a different conversation. That was in response to Bugmaster. However I will address it with you. The point I was making there is that evidence CAN show that something which we once believed to be beyond our comprehension, is within our comprehension. Of you define “supernatural” as “currently lying beyond our comprehension,” then something which was once beyond human comprehension, like lightning, went from being supernatural to natural once we understood it. Of course it might remain supernatural to some who still don’t understand it. If you define “supernatural” as “perpetually beyond our comprehension,” then nothing is truly supernatural because hypothetically everything is comprehensible even if we never actually do so.
    .
    That might be WHY no amount of evidence can prove that something is supernatural by definition. Once you provide hard evidence and comprehend it, the principle ceases to be beyond your comprehension.
    That’s where the beef with creationists may come up. They want things to be mysterious and “supernatural,” yet scientific and within our capacity to measure at the same time. That’s the contradiction / oxymoron. It is trying to be ‘A’ and ‘not A’ at the same time.
    .
    Yes as a good scientist one can entertain that even with an epsilon of zero that we are not completely certain that a situation will not occur. Maybe something can remain “supernatural” despite having evidence to prove it which by definition makes it no longer supernatural. Matt DOES entertain and talk about that, just not in very practical terms.

  60. Monocle Smile says

    @EL

    The theist sees Matt demean the theist’s position because they have no evidence, then ask for evidence, and then not more than a few seconds later also say that no amount of evidence could ever show that the theist is right. That’s the clear unequivocable end result of what Matt is saying, and I want him to knock it off. It makes the rest of us atheists look bad

    I find this to be dishonest and ridiculous on multiple levels. Nothing Matt Dillahunty says could ever make another atheist look bad, and fuck the theists who think otherwise.

    Matt has often claimed in no uncertain terms that no amount of evidence could ever possibly show that something supernatural exists

    That is not what Matt has said, and I think you need to do a better job listening.

  61. masqueofred says

    Matt, you said that no one argues for a god that defies the rules of logic. I can tell you that unfortunately you are wrong. I had multiple conversations with someone who not only said that God could create a boulder so heavy that He couldn’t lift it, but that then He would lift it. In another conversation, they said that God could both exist and not exist, because He was all-powerful, and therefore could defy logic.

    It was incredibly frustrating.

  62. kudlak says

    For the “energy cannot be destroyed” argument to work you’d have to claim that the “soul energy” escapes the body after death, unchanged and intact, which is something we just don’t see in nature, is it? I mean, it’s not like our body heat just jumps out of our corpses as a whole and remains that way instead of gradually leaching into the atmosphere after we die, merely adding to the ambient temperature.

    I guess they could argue that the consciousness is transformed from the chemical interactions within the brain into an incorporeal soul, but that isn’t how theists typically view the soul, is it? When pressed, most would still say that what they call their soul isn’t just the chemical interactions within their brains they experience presently, and that still leaves you with the problem of how that energy manages to remain together without a container.

    So, to say that you have soul energy that leaves your body and remains intact is an appeal to the supernatural, not science, and it becomes even more so when you claim that this disembodied energy can possess intelligence. Where in nature do we have any examples of intelligence without a brain?

  63. John Iacoletti says

    I have a couple of problems with the above reasoning.

    – I disagree that “I don’t know” is equivalent to 50% confidence. I think “I don’t know” is anything other than 100%.

    – In legal terms, a person can be “not guilty” and “not innocent” at the same time. His guilt is not proven and his innocence is not proven. That’s where we are with the statements “some god exists” and “no god exists”.

  64. ironchops says

    @47 & 48
    ac·quit – free (someone) from a criminal charge by a verdict of not guilty.
    OJ was acquitted for the murder but later found guilty for their wrongful death. How can that be? The verdict of not guilty does not equal innocent.

    @69
    Does supernatural really mean natural but we don’t fully understand yet? Can there be life that only exists as energy?

  65. Chris says

    @Matt (comment 62)
    I’m addressing your ‘objections’ the best I can, but you’re not giving me much to work with. Your main objection it seems is that I haven’t demonstrated my claims, but I believe I have.
    .
    I began by providing the counter arguments, and highlighting the fallacy of understated evidence.
    .
    But you don’t like the fallacy because you think you can’t assess the probabilities.
    .
    Then I provided the context in which the fallacy is valid, and a means to assess the probabilities: epistemic probability/inductive arguments
    .
    But you don’t like epistemic probability because you don’t think you can assess the probabilities…even though epistemic probability does exactly that: assess the probabilities.
    .
    I don’t intend to use ‘incredulous stare’ as an insult, I think there are occasions where it’s warranted. However, it seems your position is to assess theistic claims as uncharitably as possible; it’s magic, you can’t assess the probabilities, it can predict anything, etc.
    .
    I believe this can not only be perceived as attacking a straw-man, but it can undermine your other objections to theistic arguments. You agreed with the objections I listed to consciousness, but on what grounds are the objections valid? We take the theory of evolution as the best explanation we have for the evidence, well, that’s a comparative statement. On what grounds are we making that comparison, what if a Christian objects by saying: “that’s based on the assumption that the probability of evolution can be known”?
    .
    I hope this provides more clarity to what my position is.

  66. frankgturner says

    @ masqueofred #68
    I have met that type of person and that type of argument has been discussed on other boards here. The idea is that some want their God to be tangible and measurable, and hence testable and falsifiable so that they can claim it is real, yet at the same time make it unfalsifiable so that it cannot be debunked. That winds up being impractical and as Matt and I have discussed, it becomes sort of like that episode of The Twilight Zone where the guy gambled and never lost. FYI, he was being tortured with boredom for eternity. Essentially it is a desire for a game that has all the benefits of accomplishment (being able to loose) but which you can’t actually loose, which is a game of solitare.
    .
    The response I have for that type of individual is that by defying the laws of logic, their personal deity only exists in their mind. While it may help them cope (which I am fine with if there are no alternatives) it doesn’t help ME to cope, it impairs me. So by stating that their God functions in that fashion, they have told me exactly WHY I cannot have a personal relationship with it. I said something to that effect to a believer and his last significant words about religion that he has ever to me where, “well I can understand that and maybe you are not meant to have a personal relationship with God.”

  67. kudlak says

    @ironchops #71
    There are many examples of phenomena which were presumed to have a supernatural origin until science discovered the natural one, and there are still examples of phenomena which are not as yet explainable by science with a few that may be beyond our ability to ever investigate completely, but there are no examples of any phenomena which science has determined must have a “supernatural” cause.

    Any claim that the theorized “supernatural” constitutes the best explanation for any of these gaps in our knowledge would require a leap of faith that a natural explanation will never be found to disprove that assumption and I highly doubt that such a thing could ever be objectively proven. There seems to be no foreseeable limit to the improvement of our testing equipment and each new great mind has the work of past genius in which to build upon. Lots of theists may have a fantasy about something appearing one day to the entire world claiming to be God, but if they were honest with themselves they would realize that there was no way of proving such a claim. Lots of us have seen enough Si-Fi to immediately suspect holograms, or even highly advanced aliens with sinister anterior motives before jumping to the God conclusion.

    Even if you could prove that a god, a ghost, or something else typically classified as being “supernatural” exists, you then have to determine whether or not that thing exists, or at least originated outside of the natural realm. It may just be that we have to extend our definition of “natural” to include such previously unproven phenomena. As with quantum physics we may even be forced to concede that there are different realms within nature where the natural laws are not the same.

    As to the question about life existing as just energy I would have to say that such a thing would require both a different kind of life and a different kind of energy then any we currently have examples of. Currently, we understand that life does require energy, but also a physical body of some sort. We also understand that energy tends to dissipate in the absence of something to contain it, but I guess there could be some kind of simple force which might keep an amount of a particular energy together, but that still wouldn’t solve the problem of intelligence. For such an entity to also have intelligence would require a complexity of interaction within the energy fields, and maybe even a variety of different energies analogous to organs, each with its own containing field, and once you think about such complexity the notion of a “pure” energy being seems less and less likely.

  68. says

    EnlightenmentLiberal says
    “Bayesian reasoning”

    It seems to me EL you are getting VERY close to arguing about how many angels dance on the head of a pin, and insisting that that argument is the only one of importance, and you are the authority on the subject. It may be a interesting philosophical point but the number of angels is of no relevance in the real world, and certainly no use on a show talking to uneducated (for the most part) theists.

    When you start asserting that you are the only one who truly understands Bayesian reasoning you are tipping over into a world of arrogance and self-aggrandizement, let alone being annoying to others.

  69. Bugmaster says

    @John Iacoletti #70:

    – I disagree that “I don’t know” is equivalent to 50% confidence. I think “I don’t know” is anything other than 100%.

    I am tempted to say that this is just a difference in semantics due to the ambiguity of the English language. The words “I don’t know” could mean, “I am not absolutely certain”, or instead, “I can’t even venture an educated guess”.

    But, on reflection, I think this is a fundamental disagreement about epistemology. What does it mean to “know” something ? Most people — Christians included — would say that you can either know something, or not know it, or just be completely undecided on the issue; those three options are your only choices. The concept of reasoning under uncertainty, and doing so all the time, is sufficiently counter-intuitive that very few people give it any thought.

    To illustrate the difference, here’s a quick word problem that I often ask to software engineers who come to interview at our company:

    “Imagine that you live in a high-crime area. On any given night, there’s a 10% chance that someone will attempt to steal your car. Your car comes equipped with a Cheaps-R-Us brand car alarm. If a car thief attempts to break into your car, there’s a 90% chance that the alarm will go off. Unfortunately, the alarm is a little sensitive; and so it may go off even when there are no thieves — e.g. when a cat runs over the car’s hood, or a truck goes by on the street, or something. The chances of this happening are 80%. You wake up one night and hear your alarm. What are the chances that a thief is breaking into your car ?”

  70. Monocle Smile says

    @Chris

    On what grounds are we making that comparison, what if a Christian objects by saying: “that’s based on the assumption that the probability of evolution can be known”?

    I think at this point you’re conflating “epistemic probability,” which appears to be related to psychology, and probability as it relates to the actual state of affairs in reality. I think your lines of thought fall apart when we have actual testable claims. A christian who objects in the manner in which you speak is talking nonsense because that crap could then equally apply to every claim and thus the conclusion is that no claim can ever be evaluated. No one should give a shit about that kind of scorched-earth apologetics.

  71. John Iacoletti says

    “Most people — Christians included — would say that you can either know something, or not know it, or just be completely undecided on the issue; those three options are your only choices.”

    Whereas I would say that there are 2 choices. You either know something or you do not. Undecided means that you don’t know it.

    Is the answer to the puzzle 32%?

  72. EnlightenmentLiberal says

    @jeanettecorlett-black

    It seems to me EL you are getting VERY close to arguing about how many angels dance on the head of a pin,

    This is not an obscure esoteric matter. This is about important, core, ever-present principles of justification, belief, and knowledge. It is not esoteric to argue about the need to consider alternatives before passing any judgment whatsoever on a claim, positive or negative, except perhaps “I have not yet considered that claim and so I am undecided”. Again, something only counts as evidence for a proposition X if the proposition X predicts that we should see that evidence and no other competing proposition predicts that we should see that same evidence. That is a bare minimum of proper empirical reasoning, and that position is being denied by John and Matt.

    When you start asserting that you are the only one who truly understands Bayesian reasoning you are tipping over into a world of arrogance and self-aggrandizement, let alone being annoying to others.

    I feel very frustrated and outraged and wronged, but I’ll try to keep my first post to you civil. I have never claimed that I am the only one who properly understands Bayesian reasoning. I have claimed that it seems very evident that Matt and John do not. Completely different claims. Your post consists of the following fallacies: strawmanning, ad-hominem. I would appreciate if you addressed my real points rather than imagined or exaggerated points, and I would appreciate if you addressed the points rather than name-call.

  73. EnlightenmentLiberal says

    @John

    – In legal terms, a person can be “not guilty” and “not innocent” at the same time. His guilt is not proven and his innocence is not proven.

    I agree.

    When someone says “I believe that he is guilty of the crime”, he is expressing his confidence that the claim “he is guilty of the crime” is true. For the sake of argument, suppose I use 90% confidence for the threshold between “hunch” and “belief”. When that person says “I believe that he is guilty of the crime”, he is expressing that his confidence that the claim is true is some number in the range [90, 100%]. Perhaps it’s actually 90%. Perhaps it’s 95%. You would need to inquire further to find out.

    When someone says “I believe that he is innocent of the crime”, he is expressing his confidence that the claim “he is innocent of the crime” is true. Again, using the same “for the sake of argument” number, this person is expressing that his confidence lies in the range [90%, 100%]. His confidence is a particular number in that range. It might be 90%. It might be 95%. He would have to share additional information for you to narrow down his number. (He could also share his number.)

    I again say that it should be blatantly obvious to simultaneously say 1- “I am not yet convinced that the person is guilty” and 2- “I am convinced that it is false that the person is innocent”. From the first statement, I can infer the person has an estimate in the range:
    0% < P(guilty) < 90%
    If the person is logically consistent, it necessarily also constrains his belief in the estimate of innocence:
    10% < P(innocence) < 100%
    From the second statement, I can infer the person has an estimate in the range:
    0% < P(innocence) < 10%

    Again, here’s the obvious contradiction:
    0% < P(innocence) < 10%
    10% < P(innocence) < 100%
    He expressed two different incompatible ranges with his two statements. When someone says “I am not yet convinced that the person is guilty”, I do learn something about his beliefs w.r.t. the claim “the person is innocent”: I learn that the person does not believe that the claim “the person is innocent” is false.

    Again, this is a basic takeaway of Bayesian reasoning, and this is another point which the show has gotten wrong for a long time.

    @John
    Second point.

    Let me try to make this short. Almost as a definitional matter, as a first approximation, evidence counts as evidence in favor of a hypothesis if-and-only-if that hypothesis predicts that the evidence should exist, and no competing hypothesis predicts that the evidence should exist. That’s the difference between actual evidence in favor of a proposition and mere non-sequitir. Something doesn’t count as evidence in favor of a proposition if its competing propositions also make the same prediction.

    The only way to make any judgment of an empirical claim is to look at what evidence supports it, and the only way to determine which evidence supports the claim is to compare the fit of the evidence to alternative claims.

    Again, this is a basic takeaway of Bayesian reasoning, and again, this is another point which the show has gotten wrong for a long time.

  74. Bugmaster says

    @John Iacoletti #79:

    Is the answer to the puzzle 32%?

    No, and I could tell you what the answer is — should I do that ? I don’t want to accidentally spoil it for you. Anyway, this puzzle is a pretty standard example of the Bayes Rule, which is why I ask it of people who apply for a scientific or technical positions. And here’s why:

    Whereas I would say that there are 2 choices. You either know something or you do not. Undecided means that you don’t know it.

    As a general principle in life, this system actually works pretty well. However, it completely breaks down whenever you need to make any kind of a complex and important decision; because, in the real world, you can rarely obtain 100% confidence in anything. This is especially obvious in the world of science, where every measurement you take has some amount of uncertainty built in.

    Imagine, for example, that you are trying to buy a new car, so of course you want to get the best deal possible; for simplicity, let’s imagine that you yourself know very little about cars. So, which car should you buy ? Well, the Car and Driver magazine says that the Maibatsu Monstrosity is the best value for the money, and that car is in your prince range. However, the blog DetroitAuto.com endorses the 6000 SUX (also in your price range) as being superior to the Maibatsu. Your buddy Frank races cars on weekends, and he says that the SUX and the Maibatsu are basically both ok.

    Now, you know for a fact that neither the magazine, nor the blog, nor your buddy are infallible. The magazine takes advertisement money, so it could be biased; the blog is based in Detroit, which is where the SUX gets made, so there could be some bias there as well. Your buddy Frank cares only about racing, but you also care about safety, so his advice may not be entirely spot-on.

    If you apply your current technique to this matter, then you’d be forced to say, “I don’t know which car is best, I’ll just pick one at random”. However, if you allow for some uncertainty in your reasoning, you can come up with an answer such as, “Based on all the evidence, it is 85% likely that the SUX is superior”. This is not as good as being 100% certain, but still a lot better than nothing ! In this case, it would be irrational to buy the Maibatsu.

    By the way, a guy named Nate Silver applied the same exact techniques to politics, to predict US election results with a hitherto unprecedented degree of accuracy.

  75. Monocle Smile says

    @EL
    I think you’re conflating “considering alternatives” with “holding a position on alternatives.” Granted, this has not been very clear, but when considering the claim that X is true, it is necessary to consider the idea that X is false, but it’s not necessary to hold a position or belief about the claim that X is false. There seems to be a confounding factor here, which is the threshold at which someone is comfortable claiming belief.

    Again, this is a basic takeaway of Bayesian reasoning, and again, this is another point which the show has gotten wrong for a long time

    “Wrong” is not the same as “unclear,” and I think you’re often too quick on the trigger to make absolute claims of wrongness.

  76. John Iacoletti says

    “When someone says “I am not yet convinced that the person is guilty”, I do learn something about his beliefs w.r.t. the claim “the person is innocent”: I learn that the person does not believe that the claim “the person is innocent” is false. ”

    That’s a fair statement. But I don’t think anyone on the show would disagree. All we are saying is that “I am not convinced that the person is guilty” does not imply “I am [somewhat] convinced that the person is innocent”.

  77. Monocle Smile says

    That’s a bit clearer. The show takes a thousand-foot view of the situation. EL is whipping out his magnifying glass. He’s not wrong in his spiel on Bayesian reasoning, but it doesn’t seem like everyone’s having the same discussion.

  78. EnlightenmentLiberal says

    @John
    For the first time in a while, fully agreed. I feel better now.

  79. EnlightenmentLiberal says

    @MS

    Granted, this has not been very clear, but when considering the claim that X is true, it is necessary to consider the idea that X is false, but it’s not necessary to hold a position or belief about the claim that X is false. There seems to be a confounding factor here, which is the threshold at which someone is comfortable claiming belief.

    It would be silly to demand that you need to have a position on the claim “X is true” before you can consider the claim “X is false”.

    However, it is also true that any position on the claim “X is true” entails some corresponding “position” on the claim “X is false”. If you’re undecided on “X is true”, then you are undecided on “X is false”. If you lean in the direction “X is true”, then you lean away from “X is false”. You cannot examine a claim in a vacuum from its competing claims, and you cannot make a claim about the truth of a proposition without also making claims about the truth or falseness of competing propositions.

    Again, I think this is deathly important w.r.t. several conversations, specifically Matt’s intrinsic methodological naturalism failure – at least before 3 weeks ago. Matt thinks that evidence can show something is natural, but argues that evidence can never show something is non-natural e.g. supernatural. I’m sorry – that’s logically inconsistent for precisely the arguments I’m making in this thread. It’s not esoteric. It’s vitally important.

    It’s closely related to the idea of falsifiability. The requirement of falsifiability is a consequence of Bayesian reasoning. For example, when Matt makes those claims, in effect he claims that “X is natural” is a claim which he can support with evidence, but which is not falsifiable. That should immediately trigger anyone’s pseudo-science alarm. Talking about examining competing hypotheses AFAIK is a generalization of the falsifiability criterion.

    And I still reject the notion that this is bird’s eye view vs arcane details. No, this is critical and fundamental. This is like step 1 of how to be rational and use science, which I hope the show wants to promote, and promote correctly. Every time I see Matt make these bad arguments, I see how he’s being in effect just as dogmatic in his unfalsifiable views, and it’s frustrating, and I know I’m not the only one who sees this. I have at least one formal debate where this point was made by the theist side quite brilliantly. We as the atheist movement have to stop doing this intelligently bankrupt move.

  80. Bugmaster says

    In general, I do believe that the main difference between theism and atheism is one of epistemology. Most atheists (myself included) use evidence as the primary mechanism for deciding what is true. As the result, we can never claim that any belief (besides some trivial ones such as “A & ~A = false”) about the real world is 100% true. We cannot say, “the Sun will come out tomorrow for sure”; the best we can do is say, “the chances of the Sun not coming out tomorrow are so small that we can just ignore that possibility”.

    Most theists, on the other hand, form their beliefs based on a variety of sources: tradition, perceived morality of the people who adhere to the belief, the expert opinion of divinely inspired authorities, personal revelation, etc. They tend to use Boolean logic when thinking about the world; either something is true, or it’s false, and if you don’t know which is which then you should just reserve all judgement.

    For this reason, it’s very difficult for theists and atheists (or, at least, for the theists and atheists of the kind that I’ve described) to agree on anything, because our core thought processes are completely incompatible. This is why theists often confuse statements like “I don’t believe in any gods” with something like “I am totally certain that no gods exist” — it’s because the first statement, which means something like “I think that P(God) < 0.01" simply cannot be expressed in the theist’s epistemology.

    Unfortunately, all of us — even the atheists ! — share a vast body of rich philosophical tradition, based entirely around Boolean logic, because probability theory didn’t even exist until about the sixteenth century, and philosophy is a lot older than that. Because of this, I disagree with EL somewhat: yes, it would be nice if everyone used the Bayes rule for everything, but it’s just not going to happen in this world we are currently living in. Well, that, and also because I don’t really see a useful difference between saying “The Sun will come out tomorrow”, and “I am 99.999999999% sure that the Sun will come out tomorrow”, most of the time that’s just pedantry.

  81. Monocle Smile says

    Matt thinks that evidence can show something is natural, but argues that evidence can never show something is non-natural e.g. supernatural. I’m sorry – that’s logically inconsistent for precisely the arguments I’m making in this thread. It’s not esoteric. It’s vitally important

    Given that he’s been backed away from that stance for even longer than 3 weeks (most notably with the modifier “currently”), I wish you’d stop harping on this. I don’t find it particularly honest to hound someone for positions that have since changed.

    However, it is also true that any position on the claim “X is true” entails some corresponding “position” on the claim “X is false”. If you’re undecided on “X is true”, then you are undecided on “X is false”. If you lean in the direction “X is true”, then you lean away from “X is false”. You cannot examine a claim in a vacuum from its competing claims, and you cannot make a claim about the truth of a proposition without also making claims about the truth or falseness of competing propositions

    Agreed, but you’re missing something…the leaning doesn’t actually matter if it doesn’t cross set thresholds and the conversation is only about those thresholds. I find that there’s a very critical difference between “examine” or “consider” and “hold position (that is across the belief threshold).” It sounds like that was John’s implication.

  82. StonedRanger says

    So Im 60 years old this year. I never heard of Bayesian reasoning until maybe a year ago on this blog. Does that mean that every decision Ive ever made was bad or wrong because I don’t think like EL or use Bayesian reasoning? That’s the message Im getting. Maybe Im too old and stupid to understand. I would never think like that, asking what if this and what if that and likelihood of percentage of correctness. I probably couldn’t even tell how or why I make decisions. I just do. Some require more thought than others, I get that. I should probably just stay out of this one.

  83. EnlightenmentLiberal says

    @StonedRanger
    Bayes equation has been around for forever, as have the application to all rational thinking.

    One reason it’s only recently making a resurgence is that calculating it formally for non-trivial cases was too-computationally expensive until modern computers.

    About the proposed question: Do I think that you have not been thinking rationally because you don’t know what Bayes equation is.

    Let me start with an analogy. Some Christians will claim that morality comes from the bible, and from the 10 commandments. Of course, this is flagrantly false, and IMHO one of the best demonstrations of this fact is that people before the bible knew that killing was wrong. I want to paint Bayesian reasoning in a similar light. All of what you know about sound reasoning is very likely true, and you didn’t need to know the name “Bayes” nor know enough math to be able to practice generally sound empirical reasoning.

    To use another analogy, before the time of formal syllogistic forms, people seemed to reason just fine about most everyday things. You didn’t need to know the rule “modus ponens” by name, and you didn’t need to know that formalism, in order to do proper reasoning.

    However, whether you like it or not, modus ponens is a part of what it means to do correct reasoning. You don’t need to know this fact to do correct reasoning, but if you want to claim that some form of reasoning is valid which happens to contradict the basic syllogistic rules of logic, then you’re shit out of luck; you would be wrong.

    Similarly, you don’t have to know the math of Bayes equation to practice proper statistical empirical reasoning, but if you want to make an argument that contradicts the math of Bayes equation, then you are shit out of luck; you would be wrong.

    So no, I don’t think that you have been doing bad reasoning on every decision because you didn’t happen to know a particular sequence of English phonemes which constitute the pronunciation of a particular English word whose origins are largely an accident. It would be just as true whether it was named Bob’s equation, or if it had no name at all. It’s a simple mathematical theorem from the basic rules of math probability theory.

    Having said all of that, I think there can be value in studying syllogistic forms and formal logical arguments. In basically exactly the same way, I also think that there can be value in studying Bayesian reasoning. It will help you codify much of the correct reasoning you are already doing, and it will help you identify places where you have been doing bad reasoning – again just like studying syllogistic forms can do.

  84. EnlightenmentLiberal says

    @MS and Matt
    I’m sorry for attributing arguments to Matt which Matt may have repudiated in the last several weeks. My apologies. I’ll refrain from attributing anything more to Matt unless he says it here, or until after I watch those episodes.

  85. Patrick67 says

    77 Bugmaster

    While I have never left a comment on this site before, I do enjoy reading the comments left here. I’m definitely agnostic when it comes to god and most likely, most people would consider my views to be atheistic. I certainly haven’t had the academic training that many of the posters here appear to have, but I have managed to keep very well read in science and technology, arts, music, and to some extent politics. But I suppose no one would ever call me a genius on any matter. I’m more of a jack of all trades and master of none. You’re puzzle intrigued me and I just thought I’d throw out my guess. Is the any chance the answer is 10%?

    My reasoning is that as stated in the first line of your puzzle, on any given night there is a 10% chance of someone attempting to steal your car. In other words, whether or not your car alarm goes off in the attempt or whether said alarm goes off for some other reason or not, has nothing to do with how likely your car will find someone trying to steal it. There will always be a 10% chance of theft happening every night no matter what.

    Like I said I’m no genius but I have a high level BS meter and the question in the puzzle is setting it off.

    If you live in a high crime area where every night you face a 10% chance of be a victim of GTA, might I suggest one or two things?
    1: Move to a new neighborhood with as much haste as possible.
    2: Purchase a 12ga shotgun and start sleeping in the back seat of your car every night.

    I may very well be wrong in my guess and missed some deep dark secret with my logic, but I just thought I’d give it a shot. I couldn’t resist the puzzle.

  86. ironchops says

    @75 Kudlak
    Ok, How about these questions: Is it possible (.00000001% chance or less) that some sort of life can exist outside of our universe? If that’s was possible then relative to ourselves would that be considered natural or supernatural? If this universe was in fact created by some intelligent being outside of our universe would that being be natural or supernatural?

    @ 93 Patrick67 & 77 Bugmaster
    I am no genius either, just trying to educate myself.
    I agree with 10% as well, only because the answer sound probable. If it were me I would set up a video camera so you can catch the cop that tries to steal your car!

    @ All
    If I am 99.999% certain that I know something then can I claim that I know it? Or does anyone really know anything for certain?

  87. corwyn says

    @Matt Dillahunty:

    Any response that attempts to argue that naturalism is actually more probable must buy into their hidden assumption that the probability of theism is somehow known.

    No it doesn’t. Let’s look at the set of possible universes. If someone is willing to stipulate that there might exist a member of that set in which an omnipotent god could create (and maintain) life, without that universe being fine-tuned, then that makes our universe more likely to be naturalistic.

    The set of universe fine tuned enough for naturalistic life to spontaneously arise, and survive is necessarily less than the number of universes where a god could create life and maintain it (since is wholly contained in it). The fact that we *are* in a universe which is fine tuned is evidence that supports a natural origin for the universe, more than a created one. You don’t need to know any actual numbers to establish this, it is just a simple application of Bayes’ Theorem.

    Thank you kindly.

  88. Hippycow says

    @EL #91:
    Jeez, man, sometimes you can be a bit verbose. I think a better and more concise analogy is the fact that a dog can catch a Frisbee without being able to solve the equations to extrapolate its position based on velocity, trajectory, gravity, drag, etc. We live by and with many mathematical and physical “laws” without even knowing them explicitly.

  89. corwyn says

    @EL

    As I said above, the prudent thing to do is adopt 50 50 odds that they are guilty or innocent, e.g. 50% confidence, e.g. “I don’t know”.

    Please don’t be on a jury in which I am the defendant. The proper prior should be the probability 1 : . Until they show evidence, it should be assumed that the police just picked someone at random. If you start by assuming that they do a good job, you will be counting that evidence twice. You are just wrong here. You write with a lot of conviction about Bayes’ theorem, but you get it wrong sometimes. This can be easily seen if one imagines the police finding 10 suspects and trying them all at the same time. If the odds are 50:50 for each of them, on average 5 will be guilty of the same crime.
    Thank You Kindly.

  90. corwyn says

    @97:
    formatting failure:
    The proper prior should be the probability 1 : number of people in the area.

  91. EnlightenmentLiberal says

    Quoting Corwyn:

    Until they show evidence, it should be assumed that the police just picked someone at random. If you start by assuming that they do a good job, you will be counting that evidence twice. You are just wrong here.

    Just before what you quoted, I said:
    Bolding added

    Let’s assume for the sake of argument that I know nothing about the probability distribution of guilt vs innocence for people that appear before a jury. As I said above, the prudent thing to do is adopt 50 50 odds that they are guilty or innocent, e.g. 50% confidence, e.g. “I don’t know”.

    “For the sake of argument” is a phrase that means “I might not actually believe that it’s true, but I’m going to pretend it’s true for just a moment in order to make some other point”. I already disclaimed that the number I picked is straight out of my ass, and it’s not the actual point I’m making.

    Further, I did say “[assuming] that I know nothing about the probability distribution of guilt vs innocence”. If I know absolutely nothing about the selection process by which someone goes in front of a jury, then 50 50 priors sounds good. As soon as I start adding known general background evidence, then things change, exactly as you say.

    For example, if I assume that the police selection process is entirely random, and the population is large, and only a very small portion of the population can be guilty of the particular crime, then the odds change to something drastically low. However, that requires knowing something about the selection process, and I specifically said that for the sake of argument I know nothing.

    I might try to examine the reasons that the police brought this particular person in front of the jury in order to determine priors, but that’s a duplication of the evidence that would go into the particular-evidence. As you do know, it doesn’t matter if you put the evidence in your priors or in your particular-evidence, as long as you’re consistent. You make this point quite rightly here:

    If you start by assuming that they do a good job, you will be counting that evidence twice. You are just wrong here.

    However, again, I think that this argument only makes sense assuming I took into account other background knowledge to weigh it against 50 50, such as the size of the population and the likely size of the subpopulation which are guilty of this particular crime. You make this clear when you say:

    This can be easily seen if one imagines the police finding 10 suspects and trying them all at the same time. If the odds are 50:50 for each of them, on average 5 will be guilty of the same crime.

    Again, I did flatly say “[assuming] that I know nothing about the probability distribution of guilt vs innocence for people that appear before a jury”.

    A little charity of interpretation might go a long way. I know some people think that’s hypocrisy coming from me in this conversation, but I politely disagree.

  92. EnlightenmentLiberal says

    PS:
    I still don’t know offhand how I might determine this number, and it would be a terrible problem if I was on a jury. I still don’t know the right answer. I’m not sure offhand of the proper way to resolve the math and conditional probability. I don’t know how I would model it.

    How would you model it Corwyn? How would you decide?

    It seems an obvious fact that picking someone at random from the US population means that the likelihood of guilt for a particular crime like murder is near nil. It also seems obvious to me that this is not the proper prior to start with when examining the particular evidence in a murder trial.

    I suspect that I should use a better reference class. In most cases, I’m not examining the scenario of a random member of the population. Instead, in most cases I’m judging a family memory, friend, or acquaintance, which drastically changes the odds from the above. Even for the murder by a stranger which is IIRC rarer, it would suffice for the police to produce evidence that he was in town that day to again drastically reduce the odds from “1 vs the entire US population” to “1 vs the population of the town”. So, the police could produce what could be almost trivial evidence in order to change the reference class from “any human on the planet” to a much smaller reference class, and I think that’s how you solve the problem.

    I think that’s right. Does that sound right Corwyn?

    PS:
    In nearly all mundane everyday cases, I also do not endorse using the actual math of Bayes equation. It’s tedious, as we see now. I just endorse understanding it to better understand what constitutes good and bad reasoning. Again, example: If someone claims that evidence can support proposition X, but evidence cannot convince him of proposition not X, then that’s an unfalsifiable proposition, and there’s some logic and reason fail going on.

  93. corwyn says

    @100 EL:
    The proper reference class is ‘everyone who could have committed the crime’. As a properly ignorant juror, you should start with the assumption that that is *everyone*. It is the job of the prosecution to whittle that down to one.

    it would suffice for the police to produce evidence that he was in town that day to again drastically reduce the odds from “1 vs the entire US population” to “1 vs the population of the town”.

    No, they would need to show that no one else from ‘entire US population’ was also in town (and that being in town was a requirement). You couldn’t make an exception for just the suspect. That is probably usually a small percentage though. But imagine mardi gras in NOLA as a counter example.

  94. favog says

    Re the car alarm question: I’m assuming that the over sensitive car alarm always goes off when the car is burglarized, so there’s a 10% chance right there that the alarm goes off. There’s also an 80% chance that it goes off any way, so your car alarm goes off 90% of the nights (and there’s a 100% chance all your neighbors hate you). Since your car alarm went off, it’s one of those nine out of ten nights, but there’s only a 1 out of 9 chance it’s an actual crime.

  95. EnlightenmentLiberal says

    @corwyn

    No, they would need to show that no one else from ‘entire US population’ was also in town (and that being in town was a requirement). You couldn’t make an exception for just the suspect. That is probably usually a small percentage though. But imagine mardi gras in NOLA as a counter example.

    Meh? AFAIK, no event involves a significant portion of the US population moving into a single city at the same time. It doesn’t matter if Mardi Gras involves people coming from all over the country or the world. Only the size of the reference class matters for this purpose, e.g. the number of people who could have done it, e.g. the population of the town at the time including visitors for Mardi Gras.

    Of course, to make use of this information, we also need evidence that puts the suspect in town at the time, e.g. evidence which places the suspect inside the reference class.

    Of course, I think there are other relevant factors too, such as the reference class of people with a motive.

    PS: I still meant what I said when I said (paraphrase) “If I know nothing”. Knowing something about the size of the population is knowing something.

  96. corwyn says

    @57 Bugmaster:

    An Bayesian atheist’s prior probability for a god’s existence is somewhere close to epsilon; but a believer’s prior is somewhere closer to (1-epsilon).

    That is just nonsense. If that were true no one would feel the need to go to church.
    Here is how you find a proper prior for the existence of a god. Ask someone what their prior is for the probability of the existence of X. Don’t define X. Once you have the prior than you can start adding attributes.

  97. corwyn says

    @103 EL:

    no event involves a significant portion of the US population moving into a single city at the same time.

    That is irrelevant, the question is which of them *could have* been in the city at the time. Ruling out anyone requires specific evidence that they *weren’t* there. The sum of the probabilities of guilt for each person on the planet = 1. Absent any information, that probability gets spread evenly, 1/pop, by the principle of indifference.

  98. EnlightenmentLiberal says

    @corwyn
    Not true. Think about it as a simple “select 1 ball”.

    Suppose there are 6 billion balls in bucket X (number of people in the world). Suppose that none of those balls are red (guilty). Suppose someone selects at random 5 millions balls from bucket X and puts them in bucket Y (population of a city). Then he paints red one of the balls in bucket Y (putting the ball back into bucket Y). Then he selects at random one of the balls from bucket Y. What are the odds that the selected ball is red? About 1 / 5 million, not 1 / 6 billion. I don’t need to know which members of the larger population make up the smaller population in order to make this calculation.

  99. corwyn says

    @EL:

    I did say “[assuming] that I know nothing about the probability distribution of guilt vs innocence”.

    The trouble with that is that it is improper Bayesian analysis to a courtroom situation. If you did know that information, applying it would be fallacious. Let’s say 60% of all defendants are found guilty, does that increase the probability that the defendant is the one who did it? What about in a country where they use torture until they get a confession and *all* defendant are convicted, does that make it more or less likely that the defendant is the one who did it?

  100. corwyn says

    @106 EL:
    You are completely misunderstanding the problem here. You are claiming knowledge (“Suppose someone selects at random 5 millions balls from bucket X and puts them in bucket Y (population of a city).”) that you don’t have (yet).

    Let’s look at that example correctly. There are 7 Billion balls in a bucket. At the time of the crime 6 million of them were in the city bucket, 1 committed the crime and was painted red. Then those balls were dumped back in the huge bucket. In order to narrow it down to 5 million you need to know which balls were in the city bucket at the time (not which balls are currently in the city bucket). Knowing that is evidence. If you are bringing evidence pertinent to the case into the courtroom as a juror, you are violating the ethics thereof.

  101. EnlightenmentLiberal says

    @corwyn

    You are completely misunderstanding the problem here. You are claiming knowledge (“Suppose someone selects at random 5 millions balls from bucket X and puts them in bucket Y (population of a city).”) that you don’t have (yet).

    Quoting me:

    I suspect that I should use a better reference class. In most cases, I’m not examining the scenario of a random member of the population. Instead, in most cases I’m judging a family memory, friend, or acquaintance, which drastically changes the odds from the above. Even for the murder by a stranger which is IIRC rarer, it would suffice for the police to produce evidence that he was in town that day to again drastically reduce the odds from “1 vs the entire US population” to “1 vs the population of the town”.

    Of course, to make use of this information, we also need evidence that puts the suspect in town at the time, e.g. evidence which places the suspect inside the reference class.

    I don’t know how I could be clearer than this.

    I think you should simmer down a little, stop acting self righteously, and maybe we can have a productive conversation.

  102. says

    @Bugmaster #77
    Although I have read up on Bayesian reasoning, I cannot say I understand it well enough to apply it to your challenge just yet. I would like to take a shot anyway. It seems to me that, given the alarm IS going off, and there is a stated 80% chance of it being set off by a non-thief, then there must be a 20% chance of it being a thief.

  103. Bugmaster says

    Regarding the car alarm puzzle / math problem:

    As I said, it’s a question about the application of the Bayes rule. Let me give you guys a hint:

    Let’s say that we have no car alarms of any kind. In this case, the answer is pretty straightforward: on any given night, you chances of being GTAed are 10%, period.

    Now, let’s imagine that you do have a car alarm, but it’s broken, and it tends to go off randomly for no reason whatsoever: the presence or absence of the thief in no way influences the broken alarm. In this case, hearing the alarm gives you no new information whatsoever, so your confidence in being GTAed is still 10%.

    In my original example, the alarm is pretty sensitive: it will go off in the presence of the thief 90% of the time, but there’s also an 80% chance that it will react to a cat, or a truck, etc. But now, imagine that the alarm has a “sensitivity” knob, and we can dial it down. When we do so, the alarm will go off 50% of the time if there’s a thief (down from 90%) — but only 1% of the time if there’s a cat, a truck, a gust of wind, etc. (way, way down from 80%). If you hear the alarm, will your confidence in being GTAed under this new setting be higher, lower, or the same, as compared to the old setting ?

  104. Bugmaster says

    I should mention that none of the answers (10%, 20%, 60%) I’ve seen so far are correct, though some are close (my apologies if you posted the correct answer and I missed it). But this is not a trick puzzle; there’s a definitive numerical answer.

    Obviously, in the real world, you would want to move out of the high crime area, buy a shotgun, form a paramilitary neighbourhood watch, etc., but these are also not the right answers to this artificially designed math problem 🙂

  105. corwyn says

    @109 EL:

    I think you should simmer down a little, stop acting self righteously, and maybe we can have a productive conversation.</blockquote

    Bwahahaha. I was purposefully phrasing my comments *exactly* like yours. You know understand what people were feeling when they were complaining about your comments.

    You're Welcome.

  106. frankgturner says

    @ corwynn #113
    Wow, you beat me to it. Bravo, it did not even occur to me that you did intentionally when I read your comments. (Golf clap).

  107. corwyn says

    @112:
    Since I am aware of Bayes’ theorem, I was avoiding spoiling.
    11%.
    The alarm is not very discerning and it adds only a little information to the existing prior probability. The likelihood ratio is 90/80. Since this isn’t very large we shouldn’t expect much improvement.
    50/1 would be much better of course. I will leave that to someone else.

  108. Bugmaster says

    *** Spoiler alert — this is your last chance to avoid Bayes rule spoilers ! ***

    Corwyn is correct. In fact, the 50%/1% sensitivity setting will give us a confidence of about 85%.

    I can explain the math, if anyone is interested, but I’d like to point out the implications of the second scenario. We have a somewhat rare (10%) event, and we have a somewhat insensitive detector (50% true positive rate), which is thankfully quite selective (1% false positive rate). In this scenario, we can’t really say we “know” anything; not in the same way that we “know” that the Sun will come out tomorrow. And yet, even given our wildly uncertain estimates, we can get to 85% certainty that our car is being stolen. This is much better than flipping a coin, and it’s good enough to start forming real policies based on this information (for example, assuming that Patrick67 is a seasoned warrior, he would be fully justified in running outside with his 12-gauge when he hears the alarm).

    And our situation gets even better if we buy multiple car alarms, or if we figure out some other parallel method of detecting thieves — even if that method is likewise uncertain.

    This is why it is, IMO, very important to stop thinking in terms of traditional Boolean logic (as John put it, “You either know something or you do not. Undecided means that you don’t know it.”), and switch to thinking in terms of probabilities. True, you have to abandon absolute certainty if you do that — but in exchange, you gain the ability to reliably understand the world and even predict the future, despite being a limited human armed with wildly unreliable senses. The tradeoff is well worth it.

  109. says

    @Bugmaster #116

    I would like to see the math, if you have the time. I guess I misinterpreted the statement “The chances of this happening are 80%.” to mean that, when an alarm goes off, 80% of the time it is a false alarm. Which seems to make the other numbers irrelevant.

  110. Bugmaster says

    Ok, here’s the math. But first, some definitions:

    This scenario is basically all about two events: the sound of the car alarm, and the presence (or absence) of the thief. Let’s use “S” to mean, “the car alarm is sounding”; similarly, let’s use “T” to mean, “a thief is stealing the car”. The probability of T happening is 10%, or 0.1 (since probabilities are usually expressed as numbers between 0 and 1). We will use the notation “P(T) = 0.1” to express this idea. The thief is either there or he’s not; since P(T) = 0.1, P(~T) = (1-0.1) = 0.9. This is a shorter way of saying, “there is a 90% chance that the thief is not present”.

    If there’s a thief present, there’s a 90% chance that our alarm will sound. We can express this as “P(S|T) = 0.9”, which we can read as “The probability of Sound given Thief is 0.9”. This means that, the other 10% of the time, a clever thief can sneak right by our alarm without setting it off. Unfortunately, our alarm is on a hair trigger, and P(S|~T) = 0.8. This means that, the other 20% of the time, random cats and trucks and such will avoid setting off the alarm.

    Just as a warm-up exercise, what are the chances that our alarm will go off, for any reason ? Well, one reason the alarm could go of is due to the presence of the thief. If the thief was always there (i.e. if P(T) = 1), then the alarm would have a 90% chance of going off. But the thief is only there 10% of the time; the other 90% of the time he’s not there, but there’s an 80% chance that the alarm would go off anyway. Thus, P(S) = P(S|T)*P(T) + P(S|~T)*P(~T). Plugging in the numbers, we get P(S) = 0.9*0.1 + 0.8*0.9 = 0.81.

    This tells us that there’s an 81% chance on any given day that the alarm would go off, but what we really want to know is, “given that the alarm is going off, what are the chances that it’s because of a thief, and not because of some random cat ?” That is, we need to know P(T|S).

    The Bayes theorem states that P(T|S) = P(S|T)P(T) / P(S). In English, this means, “Given that we are hearing the alarm sound, the probability of the thief being there is a ratio of true positives to all positives”, or, to put it another way, “the ratio of all the reasons for the alarm that we care about, divided by all the reasons the alarm might go off, period”.

    Plugging the numbers into the equation, we get: P(T|S) = (0.9 * 0.1) / 0.81 = 0.111111, or about 11%.

    But what if we dialed down the sensitivity ? Well, in this case, P(S|T) becomes 0.5, but P(S|~T) goes all the way down to 0.01. P(T) doesn’t change, it’s still 0.1. Thus, P(S) becomes 0.5*0.1 + 0.01*0.9 = 0.059. P(T|S) then becomes 0.5*0.1 / 0.059 = 0.8474, or about 85%. Our alarm is dramatically less likely to go off; but when it does, we can be fairly certain that it’s for a good reason.

  111. says

    @Bugmaster #118

    Thanks! That helps a lot. I recently read Daniel Kahneman’s book “Thinking: Fast and Slow” where, in chapter 16, he talks about how Bayesian reasoning is not intuitive. He gives an example very similar to your alarm scenario, but he does not show the math. I put his numbers into your equations and I got pretty much the same answer he gave in the book. I feel like I have learned something useful.

  112. Bugmaster says

    @Carol Sperling:
    Glad I could help. By the way, the really interesting stuff starts to happen when you start dealing with Bayesian belief networks, where pieces of evidence can support multiple conclusions, and those conclusions can act as evidence for other conclusions, etc. All of this stuff is not just of academic or philosophical interest; Bayesian networks are practical tools that allow many of our modern technologies (e.g. search, speech recognition, spam filtering, gene discovery, just to name a few) to function.

  113. Bugmaster says

    @Carol Sperling #118:
    Er, but I should add that these are not my equations, these are Thomas Bayes’s equations 🙂

  114. John Iacoletti says

    For some reason when you first posed the question, I must have totally missed the statement that says, “On any given night, there’s a 10% chance that someone will attempt to steal your car”. Not that I would have gotten the right answer anyway…

  115. favog says

    Bugmaster, it appears you did miss post #102, Probably because I wrote it in the format of “1 out of 9” instead of a percentile.

  116. EnlightenmentLiberal says

    @corwyn
    Thus far, there appears to be a difference between you and me. I’ve been painfully honest in these discussions, and I have not been purposefully a jackass to make a point. I doubt your honesty and your intentions. In other words, it seems that you just admitting to trolling.

  117. EnlightenmentLiberal says

    @Lots of people
    I do not know why several people here have decided it’s ok to troll me, and why others have cheered on people trolling me.

    I do not know why people think I’m making esoteric claims that don’t matter when several people in this thread have been clearly wrong assertions which I’ve been trying to address.

    I do not know why people think I’ve been acting self righteously – whatever that means. It sounds like a complaint from a Christian who already knows they have lost and will ad hominem rather than attack points. I still believe I am right on all of my points in the thread (give or take a very minor mistake I’ve overlooked).

    I have been very badly mistreated in this thread by several people I expect better from, and I don’t know why they have acted so badly.

  118. kudlak says

    @ironchops
    I think that your key phrase has to be “relative to”. Even in a multiverse, all things would be native and therefore natural to the system as a whole, but unnatural from one universe to another. Things foreign to a particular universe may appear “supernatural” to anyone believing that theirs is the only universe, but for something to be truly supernatural to anyone knowing they live in a multiverse wouldn’t that thing have to somehow be foreign to the entire system?

    As to what we can actually claim to know, doesn’t the “brain in a vat” argument negate anyone ever truly knowing anything for certain? I have no way of really knowing whether I’m actually here, typing away while also smelling my lunch cooking, or just a SIM running on somebody else’s computer. In a sense you’d have to be truly omniscient to know that what you are experiencing and believe that you “know” is absolutely true to zero percent chance of being wrong.

    Not that I believe actual omniscience is possible to prove to oneself. Christian God, for example, if he exists may believe that he is completely omniscient in all things based on everything he knows. He may have been the only thing that he knew existed up until he created more, but how could he ever be sure that something, somewhere doesn’t exist outside of what he can perceive and remember, or even that nothing ever escapes his notice within his own creation? He wouldn’t know what he doesn’t know, as they say, and that would also include his own limits, ruling out his ability to truly claim to be omni-anything, right?

  119. Bugmaster says

    @favog #123:
    Oops, I totally missed your comment, sorry about that.

    @EnlightenmentLiberal #125:
    I am pretty much on your side in this debate, but I gotta say, you do come off as self-righteous sometimes. This is entirely a matter of tone (i.e. it’s the dreaded Tone Argument), not of substance. The facts and concepts you are presenting are all correct, but you tend to associate subjective value judgements with them, and most humans tend to get fairly defensive when they see you doing that. Ironically, if everyone used perfect Bayesian reasoning, you wouldn’t have this problem — but then, we wouldn’t be having this debate in the first place.

  120. Monocle Smile says

    @Bugmaster
    That’s pretty accurate.

    @EL
    After re-reading Matt’s first direct response to you, I’m more on your side, although I know what he’s getting at. When someone offers up a claim, I’m not obliged to disclose my leanings; only the final evaluation (accept or reject).

    IMHO, this plays nicely into my positivism that if you do not have some (observable) alternative, then the claim itself is not well formed and the truthness / falseness is not defined for the proposition. This is represented in Bayes equation thus: If you have no alternative hypothesis, then Bayes equation is inapplicable.

    I am firmly in agreement here, and this is actually a point of contention between me and most of the AXP hosts. Tracie touches on this sometimes, but Martin hits theists harder on this point (by asking what a universe without a god would look like, which is an alternate hypothesis) than anyone, though he rarely gets an answer, and I wish the other hosts would take similar tactics. Most of the time, when a god claim comes up, Bayes is hard to apply because the god is ill-defined and we don’t even know what “evidence” would look like.

  121. says

    @EL # 17
    “Richard defines supernatural as any irreducibly mental substance – whatever that means. I think the definition is logically inconsistent and incoherent because the concept of “mind” is properly defined in terms of some sort of reduction. (For that point, please see the work of Daniel Dennett.) Thus IMHO Richard’s entire argument breaks down.”

    I LIKE Carrier’s definition for just that reason. By boiling it down to the property that so-called supernatural entities supposedly have in common (mind without matter), it strips off the emperor’s clothes, and illuminates the inherent contradiction.

  122. Hippycow says

    @Carol #129:
    Yeah, that is a kind of interesting way to look at the word. Anything that is physical is natural, right? So, we can make up a word like “supernatural” and imagine things that are “supernatural” but those “things” only “exist” in our minds. Maybe the mind is supernatural; while it only exists with a brain, it is not the same thing as the brain. Brains can exist without minds, it would seem. I guess “mind” would be something like a “self aware consciousness” that is cohesive to some extent over time, though there are times (sleeping and other) when we are not necessarily self aware.

    I guess if we get to the point were computers can thoroughly and convincingly pass the Turing Test, then we get into the realm where we don’t know if a computer has a mind. But if it does seem to have a mind and it thinks it has a mind (or at least can convincingly argue that it does), then we have minds that are unnatural and supernatural.

    However, that’s philophy, while religion is only philosophy for children. In the context of religious discussion, the most honest definition of “supernatural” is simply “make believe.”

  123. corwyn says

    @124 EL:

    Nope. You were *wrong* about how Bayes law should be applied. I pointed that out.

    Given that you are so dogmatic about how you are right and other people were wrong that you were turning people off the idea of Bayes theorem, I killed two birds with one stone by pointing out your error, in your own idiom. If you don’t like it, the solution is in the mirror.

    And in fact, you asked for this help. You said that you were having trouble convincing people. Be a good bayesian, admit your mistake, modify your writing patterns, and be more helpful is actually convincing people that bayes theorem is important. Or, behave like a caller who has had their pet argument crushed on AXP.

  124. favog says

    We’re cool, Bugmaster. Like I say, I assume that you scanned looking for “%” and I didn’t give you one.

  125. Monocle Smile says

    @Hippycow
    I get your meaning, but I’m very leery about categorizing the mind as a “thing.” The mind is what the brain does, it’s not something that exists in the brain.

  126. corwyn says

    @130:

    Are you saying that a computer mind would be ‘supernatural’?

    Also be careful about causing confusion between ‘natural’, not man-made, and ‘natural’ of this universe.

  127. says

    @Hippycow
    ” The mind is what the brain does, it’s not something that exists in the brain.”

    YES ! Thank you! Mind is a process, not a thing. I have the same issue with ID people who think Design is a property of an object or organism. It is not. Design is a process that results in a particular configuration of atoms. It is part of the history of an object, not an inherent property of the result.

    This is a failure of our language and culture.

  128. Bugmaster says

    @corwyn #134, Hippycow #130:
    See, this is exactly why I personally tend to take a sort of hard-line stance on the Turing Test and its implications.

    If I am able to hold interesting conversations with some entity (or, rather, conversations that are at least as interesting as my conversations with flesh and blood people), then it doesn’t matter to me whether that entity possesses a natural mind, a supernatural mind, an unnatural mind, a biological consciousness, a sophisticated algorithm that is completely indistinguishable from consciousness but is nevertheless not conscious, or whatever. None of that stuff matters because, at the end of the day, all I’m interested in is talking to people. If I can talk to it, it’s a person. Once you abandon evidence (here Bayesian reasoning comes into play again) and start splitting metaphysical hairs, you inevitably end up at some sort of a solipsism that is not useful to anyone (except maybe to some philosopher in search of tenure, I don’t know).

  129. Narf says

    @14 – Hippycow

    Anyway, don’t constants just come about as an artifact of our units of measure?

    Sorry about reaching way back in the comment stack here. I don’t seem to be getting the new post notifications, for some reason, and I’ve been busy so I went and checked just now to see what was up with the post from this last show.

    What we were talking about, a post or two back, was mathematical constants. Those are artifacts of the units of measure that we’ve arbitrarily assigned to various aspects of reality, yes. There are then physical constants, though, which represent the relationship between various aspects of reality, such as the speed of light in a vacuum; the relative strengths of the gravitational, electromagnetic, strong, and weak forces; the differences between various types of baryonic matter, and a few other physical constants. Those physical constants are expressed with units, so there will also be mathematical constants involved, but the physical constants lie underneath all of the math.

    A competent delivery of the fine tuning argument involves these physical constants. There are plenty of issues with the argument, even if it’s delivered competently, but what’s funny as hell is that lots of Christian apologists don’t get the argument right. I’ve seen Ray Comfort’s version of the fine-tuning argument, and more recently Carl Gallups, who is a vapid intellect about on the same level as Ray. They don’t even list physical constants, instead presenting the Earth’s atmosphere, our distance from Sol, the moon, how well we fit into our biosphere, and similar temporary things as evidence of fine-tuning.

  130. EnlightenmentLiberal says

    @corwyn

    Nope. You were *wrong* about how Bayes law should be applied. I pointed that out.

    Orly? Which one? The part where I said “for the sake of argument” to disclaim that point, and still on this point I was right and you are wrong – you seemingly didn’t notice where I said (paraphase) “suppose I know nothing about the selection process”? Or are you talking about the other point you got wrong, where you said that I would need evidence to restrict the population of interest to the size of the city, when I already explicitly made that point no less than twice previously?

    You’re not engaging in god faith corwyn. You’re not being intellectually honest either. You think I’m being an asshole, and you think you can teach me a lesson by being a bigger asshole, or something. You’re not out to correct me or to engage in a constructive dialogue.

    I’ll try to be the first to admit when I make an error, but I don’t see you identifying one yet. I just see a lot of hot air from someone upset. Upset maybe because I’m arguing that your heroes from a public access television show are wrong. I honestly don’t know, and I’m starting to not care either.

    I’m not interesting in you teaching me a lesson about humility. This is me letting you know that you’re wasting your time. I actually feel a little sorry for you too.

  131. Narf says

    @36 – ironchops

    I need some help here on this Fine tuning argument thing.
    So far we (the people (animals) of planet earth) have found no concrete evidence of life anywhere except here (earth). Using math we have determined some sort of probability that life could occur out there somewhere but that’s all we have for now as far as I know.

    We’ve detected organic molecules in plenty of other places, but as far as I know, we haven’t detected actual microbes yet.

    Question 1. How can the universe be fine tuned without life popping up all just any ole place.

    This is one of the problems with the way that theists tend to present the argument. They’ll argue that the universe was created specifically with the intention of providing us with a place to live, here on Earth, and they’ll ignore the fact that the state of the universe makes it almost certain that some sort of life is common all over the place, both in our galaxy and in others.

    One of the biggest mistakes you can make about Christian apologists is in thinking that they’re actually interested in discovering anything about reality. They’re pulling these arguments out of their ass or cobbling them together from stray bits of information that they overheard from the Discovery Channel playing in the background at the bar they happened to be in.

    Some are more honest about this than others. William Lane Craig is a thoroughly dishonest person, but he at least had the balls to state up front, in his most popular book, that reason is to be placed subservient to the authority of the Bible, and he’ll ignore anything that contradicts what the Holy Spirit told him is true.

    Question 2. Isn’t the universe constantly changing (tuning)?

    This is one of the things I just addressed in an aside to my response to Hippycow, which I wrote before reading this comment of yours. There are a lot of shitty apologists out there — probably 80+% — who don’t even know how to properly make the fine-tuning argument.

    A proper fine-tuning argument refers to physical constants which are unchanging, as far as we know. Most apologists throw in — or use exclusively — a bunch of temporary features of our planet, which were drastically different in the past, when the first organisms were developing here on Earth. Some exceptionally ignorant/dishonest apologists get even worse than that.

    Kent Hovind presents the high-oxygen content of air bubbles found in amber as evidence that the prebiotic atmosphere of Earth would have caused any organic molecules to oxidize. Look over the details of that last sentence and see if you can spot what makes the claim so amazingly stupid.

    If so then Question 3. When will it become so fine tuned that we can’t live in it any more?

    Uhhhhhhhh. I’m not sure what you’re getting at here. Fine-tuning implies that something is fine-tuned for a specific purpose. If the universe was fine-tuned for us, then making it a little more perfect for us wouldn’t hurt.

  132. Narf says

    @69 – kudlak

    I guess they could argue that the consciousness is transformed from the chemical interactions within the brain into an incorporeal soul, but that isn’t how theists typically view the soul, is it? When pressed, most would still say that what they call their soul isn’t just the chemical interactions within their brains they experience presently, and that still leaves you with the problem of how that energy manages to remain together without a container.

    I think that most theists ignore the apparent situation that what brain does is ‘mind’. They see the soul as something extra-universal which is driving the brain and making it be a mind, not something that emerges from the brain.

    Essentially, any possible mechanism you could possibly throw out as a prospective vehicle for the soul will be rejected, because they know, at some level, that they have to avoid any attempts at clarity and detectability.

    So, to say that you have soul energy that leaves your body and remains intact is an appeal to the supernatural, not science, and it becomes even more so when you claim that this disembodied energy can possess intelligence.

    We aren’t talking about actual, detectable energy here. We’re talking about new-agey ‘energy’ and ‘vibrations’. You know … supernatural energy.

    Where in nature do we have any examples of intelligence without a brain?

    “God.”
    … or some similarly vapid answer. They’re not working with actual evidence here.

  133. Narf says

    @71 – ironchops

    Does supernatural really mean natural but we don’t fully understand yet?

    Not the way that religious people use the word, but that would be an honest assessment of the sorts of things that they’re talking about, when they speak of the supernatural. When a religious person talks about the supernatural, they’re usually asserting that we’ll never understand what they’re talking about, and they’ll continue to push their concept off beyond the edges of our understanding, as we learn more.

    Can there be life that only exists as energy?

    I could see any medium that can maintain a replicating pattern as being able to manifest life. That last sentence is a little fucked, but I think I got my point across.

    Damned if I can conceive the mechanics of the situation, but I could see it as a possibility.

  134. Narf says

    @70 – John Iacoletti

    – I disagree that “I don’t know” is equivalent to 50% confidence. I think “I don’t know” is anything other than 100%.

    Uhhhhhhh, whoah. Hold on there. You just turned ‘knowing something’ into a useless concept that has no representation within reality. You might want to adjust your definition a bit, if you haven’t already, further down the comment stack.

  135. Narf says

    @77 – Bugmaster

    “Imagine that you live in a high-crime area. On any given night, there’s a 10% chance that someone will attempt to steal your car. Your car comes equipped with a Cheaps-R-Us brand car alarm. If a car thief attempts to break into your car, there’s a 90% chance that the alarm will go off. Unfortunately, the alarm is a little sensitive; and so it may go off even when there are no thieves — e.g. when a cat runs over the car’s hood, or a truck goes by on the street, or something. The chances of this happening are 80%. You wake up one night and hear your alarm. What are the chances that a thief is breaking into your car ?”

    Your wording is a little vague, as you typed it here. Is that an 80% chance of a false alarm each night or an 80% chance of it going off each time a cat runs over the hood or each time a heavy-vibration truck goes by? If it’s the latter, then the damned thing is going to be going off constantly, and you’d be somewhere down well below 1%, probably. If that’s 80% nightly, then you’re looking at just over 10% chance (10.11-something%) of any given alarm being a real thief.

    I’m a network engineer, so I’m quite familiar with the high incidence of false alarms and the joys of digging through system logs.

  136. Narf says

    @93 – Patrick67

    You’re puzzle intrigued me and I just thought I’d throw out my guess. Is the any chance the answer is 10%?
    My reasoning is that as stated in the first line of your puzzle, on any given night there is a 10% chance of someone attempting to steal your car. In other words, whether or not your car alarm goes off in the attempt or whether said alarm goes off for some other reason or not, has nothing to do with how likely your car will find someone trying to steal it. There will always be a 10% chance of theft happening every night no matter what.

    Here’s a breakdown that’s fairly easy to follow:

    SPOILER SPOILER SPOILER
    … for anyone who still hasn’t figured it out.

    I’m going to assume that the chance of a false alarm is 80% per night, rather than an 80% chance of a false alarm per cat and truck. If it was 80% per cat/truck, he would have had to have given an incidence of cats and trucks.

    The incidence of a positive alarm is 9 per 100 nights. You’re going to get 10 attempted thefts per night, but the alarm will only detect 9 of those.

    You’re going to get 80 false alarms per 100 nights, since that’s what was stipulated in the problem. So, you’re going to get 89 alarms per 100 nights, and 9 of those will be from actual theft attempts. 9 / 89 = 10.1123595505618%

    There are equations that do it a little more cleanly, but I wanted to put it in simple wording.

  137. Narf says

    @94 – ironchops

    @75 Kudlak
    Ok, How about these questions: Is it possible (.00000001% chance or less) that some sort of life can exist outside of our universe? If that’s was possible then relative to ourselves would that be considered natural or supernatural? If this universe was in fact created by some intelligent being outside of our universe would that being be natural or supernatural?

    We have too many questions and too many definitions to iron out. I’d need a coherent explanation of what is outside of our universe and a demonstration of how the person providing the information knows what he knows.

    Sorting out the cluster-fuck of verbiage is secondary.

    @ 93 Patrick67 & 77 Bugmaster
    I am no genius either, just trying to educate myself.
    I agree with 10% as well, only because the answer sound probable. If it were me I would set up a video camera so you can catch the cop that tries to steal your car!

    Close, but for the wrong reasons, I suspect. I imagine that Bugmaster learns more from the process of watching the person work it out, rather than the final answer itself.

    @ All
    If I am 99.999% certain that I know something then can I claim that I know it? Or does anyone really know anything for certain?

    You switched up scopes in the middle of that. Are we talking about knowing something, or are we talking about certainty? I set my threshold for ‘knowing’ something well below five 9’s.

  138. Narf says

    Jeez, man, sometimes you can be a bit verbose. I think a better and more concise analogy is the fact that a dog can catch a Frisbee without being able to solve the equations to extrapolate its position based on velocity, trajectory, gravity, drag, etc. We live by and with many mathematical and physical “laws” without even knowing them explicitly.

    Errrrrrrr, no it doesn’t. A dog catches a Frisbee with a vague guess at the beginning and a whole lot of on-the-fly error-correction.

    I heard this in some book or other, which I’m forgetting right now. There was some nonsense about music appealing to the part of our subconscious that can do differential calculus effortlessly, allowing us to do something so unlikely as catching a thrown baseball … which is nonsense, as I said. Our mind does slapped-together bodge-jobs of most things and learns to be very good at that, most of the time.

    Or were you not making that sort of statement but rather a statement about the way that what I described can look like precision calculation, if you’re looking at it wrong and seriously overthinking things?

  139. Narf says

    @118 – Bugmaster
    I think you oversimplified a little, actually. Your number is a little high, I think because you ignored the possibility of having both a false positive and a true positive on the same night. Your equation is a little bit cranked, because of how you plugged the numbers in.

  140. Narf says

    … or to put it another way, you’re treating each night as a single, prolonged incident. You need to adjust your equations to more closely reflect reality. The final answer could be more like 10.15% or 10.2% to reflect the chance of a cat jumping on the hood of the car while a highly-competent thief was stealing it, but it wouldn’t be anywhere near 11.11111111111%, which would require a single chance per night, for both the cat/truck and the thief to simultaneously have a go at setting off the alarm.

  141. Narf says

    Correcting #145:

    The incidence of a positive alarm is 9 per 100 nights. You’re going to get 10 attempted thefts per night per 100 nights, but the alarm will only detect 9 of those.

    Had a bit of a typo there.

  142. Narf says

    There’s another point of vagueness that I had to iron out, too. When you list an 80% chance of a cat or truck setting off the alarm, I had to treat it as an 80% chance of something like that setting it off once per night. You’d have to phrase it as something setting off the alarm 4 times every 5 nights, if you wanted it to reflect reality properly, but I was able to pull that meaning from the question, for the purposes of extracting the number.

    If you wanted it to look more natural, you could say something like the alarm being set off 6 times a week, but that would make the math uglier. 7’s are annoying.

  143. frankgturner says

    @EL
    All right this is going to take me a bit but let me see if I can get through to you. I knew this guy Fred who was similar to WLC. He was of the common mental state that the moment you show even the slightest bit of doubt that you are completely wrong about an idea. There was no “I don’t know” or 50% or 60% likelihood to him mentally. You were either absolutely certain that something would occur or absolutely certain that it didn’t (pretty Boolean, absolutely true or false).
    .
    As you might have guessed he was pretty fundamentalist, though he did not argue that every part of the Bible was correct or the inerrant word of god, he did feel that certain parts were “written in such a way as to indicate absolute certainty” (mainly Genesis) and he had read the whole thing. He was very quick on the trigger to indicate that other people were “wrong” (and would say so with a very condescending accusatory voice) and to me came off as a bully (to some others as well). A lot of people just did not talk to him. Some did as his attitude seemed to appeal to them, telling me that “he seemed so confident.” Like WLC Fred put on an air of authority to protect himself from insecurities. There was an analogy I made to him about 2 people looking at a stoplight, particularly a green one, when one person is color deficient in the green region and the other is not. He was very quick to point out that the idea was “wrong” (again accusatory condescending voice) because it was biased, the two people have different perceptions. He had not even listened to the whole thing and he was already putting it down. I continued, making a point that the analogy was ABOUT differences of perception and how the bulb behind the filter in a stoplight is a white bulb and the color filter gives the regular color vision person the perception that the light is green by filtering out other wavelengths of light and while it fails to give a color deficient individual that perception as one does not detect the green wavelengths, it does give the perception that the bulb behind is white, which it is. So despite differences of perception, both are perceiving truth in the sense of factual correctness despite the point that they are perceiving something different. Obviously one is not perceiving the “intended” truth of the light being green.
    .
    You EL are very quick on the trigger to indicate when others are “wrong” (maybe not with a condescending tone as one cannot hear it through type) but the effect has been very similar on here. If Fred had been more willing to listen to the rest of the analogy or ASK when his perception may not have matched that of his listener given the recognition that he brings different perceptions and background to his understanding of the context of a situation he would have come across as less of an asshole to me (and many others). Fred didn’t seem interested in that idea though, that others might have perceptions based on their background and that the context they bring to the situation may not be incorrect. It was not surprising as he had been taught to be closed minded about his Xtianity. Many a former Xtian or Muslim, etc, finds themselves coming away from their religion that is “all based on faith” when they look at it from the perception of, “had I been born into another religion, whose argument that ‘you have to take it on faith’ is the same as the one into which I was born, would I really find what I was born with compelling?” In other words, if you put yourself in someone else’s shoes coming from THEIR context and THEIR background and listen to your own argument, would it really be all that compelling. Many are some deeply embedded in their emotional faith that they can’t see that.
    .
    Many conservatives I have noticed see the willingness to admit the possibility that oneself is wrong about an argument as a sign of weakness, that you should express absolutely certainty of your views regardless of the evidence if you are truly confident in them. They see being opened to information that could change, facts particularly, as a sign of weakness. It is a powerful political move if you are trying to get your views to be popular and have power like WLC (as Narf mentioned, hence why I see WLC as a politician instead of philosopher). What the WLC bots and many fundie Xtians don’t see is that there are those like us (maybe you, I don’t know for sure) who see an unwillingness to consider the possibility of ones viewpoints being incorrect, we see THAT as the sign of weakness and the openness to changing ones mind if the evidence points in that direction as a sign of strength.
    .
    As a result of this idea when I see someone who is too confident about their position, too quick to insist that others are wrong and too insistent that their ideas are infallible, I am highly suspicious. Matt said something on the show once about how (paraphrasing) the best debaters know their opponents views so well that they could practically make their opponents argument for them. Thats what I thought was so hilarious about his debate with StB. Despite StB trying to manipulate the audience with videos of Matt (a clear indication that he HAD listened to Matt’s arguments, which is good practice) he did not offer ANY hard evidence as to why they would be right or wrong. Matt actually recited some of StB’s arguments before offering evidence why they were wrong or even questionable. So they obviously knew each other’s arguments well. Matt probably (I don’t know) have even considered, hypothetically, the conditions under which StB might be correct and have thought about how to address them IF StB brought them up even if StB had not in the past. (In this case it would have bee mute, when hard pressed for evidence on how StB knows shit he deflects rather than giving hard evidence).
    .
    You did not make much of an effort to understand Matt before being quick on the trigger to accuse him of being wrong rather than considering the possibility that maybe you don’t understand the context in which he is speaking (like Fred and the light bulb analogy). You said,

    I also think it rather obvious that if Matt says that you can determine guilt without examining the question of innocence, then Matt is teaching people to think badly. The straightforward understanding of that claim is simply and completely wrong, and I think it rather reasonable to understand that some people will be swayed to agree with the straightforward but wrong understanding of the claim.

    I agree and disagree with you at the same time by examining context and potential missing adjectives. When considering guilt in a courtroom we do consider the burden of proof to be on the prosecution, the individuals making the positive claim and I believe that is done for a particular reason in law. The principle in law that is discussed is that we want evidence to show guilt “beyond a reasonable doubt” primarily as we would prefer (I have heard it said) that 10 obviously guilty people go free rather than 1 potentially innocent person be wrongfully convicted. So verdicts in criminal cases are restricted to “guilty” (A) and “not guilty” (~A). Evidence could support A or not support A. There is a question of innocence (B) and non innocence (~B) and I will agree that these are influenced by whether evidence supports guilt vs non guilt because it is not a vacuum. So yes if you think about it implicitly (the missing adjective) Matt is incorrect in my view. Matt did not say whether he was speaking implicitly or explicitly though. If you think about it, juries are only asked to explicitly determine if the evidence is enough to determine guilt or non guilt. They are NOT asked to explicitly determine innocence. Evidence that demonstrates innocence can most certainly be presented as this implicitly influences determination of guilt and implicit determination of guilt can be used to make an explicit verdict of guilt. However, they are not asked to explicitly declare the person “innocent” or “not innocent.”
    .
    So you see what I did here was presented the circumstances under which I think you might be wrong EL and the circumstances under which you might be right and I am not afraid of being incorrect on either side. I think that given the information I have that you are basically correct but wrong in your presentation. You came off the same way that Fred did though from my viewpoint. And when corwynn used your same phraseology back at you it appeared to piss you off. That is sort of like the Xtians who insist that since they are not Muslim they can’t be told not to draw Mohammed and that insisting that they don’t is shoving Islam down their throats yet seem oblivious to the idea that insisting that gays can’t marry and that this should be the law of the land is shoving Xtianity down the throat of others.
    .
    You may not have intended to sound dogmatic EL, but I think you did despite my agreement with your general premise and mathematics. To me and to others you came off just like Fred did to me being quick on the trigger to insist that others were wrong and you were right without evening listening to the full argument. You even admitted to not having listened to the episode yet at one point. Corwynn was not being a “bigger” asshole, he was being the same level of asshole to see if it would cast a reflection. I don’t think that you intended to come off as an asshole EL, but it certainly seems like you were and still are in many ways. Here’s a way to sound a bit more confident of your position in an honest way to me, unpack your arguments a bit, i.e.: consider the assumptions being made and present them in a clear fashion (I am going to do this in a bit regarding “natural” vs “supernatural,” which I did to some degree already but others are doing parts of this as well) and consider the possibility that you may not be defining things the same way that other people are given that (as Matt has said on the show many times) “words have meaning by consensus.”
    .
    Unpacking the gum ball argument a bit: We can see that the jar has some gum balls in it, at least 1. In determining even or oddness we know that the number is positive and non zero. We assume that the gum balls are whole and even if we saw fractions we would not count them. We can just make an agreement to only count whole gum balls if fractions are present. We assume that we are counting in decimal. We assume that the total number is finite. There might be other assumptions other than what I have listed and please feel free to add if needed.
    .
    Based upon the general rules of decimal number system and these assumptions we can state that the number is either even or odd. This system appears to be a true dichotomy so in this case we can state that even (X) and not even (~X) are diametrically opposed to odd (Y) and not odd (~Y). So the probability of X is the same as ~Y and Y as ~X, However, this is implied if not directly stated. We don’t have to explicitly consider the likelihood of oddness when we are looking at the likelihood of even vs not even.
    .
    The possibility of a god is not such a nice clean dichotomous system whose probabilities are so well implied by an understanding of the rules of the system as we have a very limited understanding of the rules of the system, the universe. Theists like to try to claim that it is, hence the whole “if you are not for me you are against me” bullshit. We recognize that there are many more if not infinite) variables to take into account. We are better informed (or at least express a better informed attitude) about the world around us. We recognize that explicit evidence that could demonstrate the probability of a god being guilty or not guilty of existing, while it does influence the implicit probability of a god being innocent or not innocent of existing, we are making explicit statements and only making explicit considerations. Making explicit claims regarding the innocence or non innocence of a god existing is a separate consideration.
    .
    You never made an effort to ask Matt in what context he was speaking or whether he was considering explicit or implicit terms. You just insisted that he was wrong, as though you knew his context better than he did. That comes across like most Xtian apologists that I have observed and it does sound like an asshole move. Is any of this getting through?

  144. frankgturner says

    Oh and something that theists and apologists do that makes them come off like suspicious assholes to me, take their opinions and subjective ideas and present them as though they were empirical fact. To me that is a bully tactic made to make oneself sound over confident and portray an air of authority. To me it says the speaker is afraid of being wrong. Couching terms, eeven objective data, with “I think,” “in my opinion,” or (among others) “I feel that” says to me that the speaker is not afraid of being incorrect and has a solid, yet flexible and open minded position. We do that some with IMHO but I have spoken about couching before.
    .
    Personally I wish we could just reach into our brains and turn off the emotion switch for a while but it is woven into our thought process.

  145. favog says

    Narf, you’re not understanding the defect of the alarm. It’s over sensitive, so it gives false positives, but it does detect all of the actual theft attempts, so it gives no false negatives. Thus, it is indeed 10 nights /(10 nights +80) nights for a 1 in 9 frequency.

  146. Narf says

    From the problem:
    “If a car thief attempts to break into your car, there’s a 90% chance that the alarm will go off.”

    That’s a 10% chance of a false negative.

  147. favog says

    I completely missed that detail! Even made an aside in the original solution I had.

  148. Narf says

    Heh. There was a bit of wonky wording in the problem and a few assumptions you had to make, as I pointed out in mine. That detail probably got lost in the other assumptions you had to make.

  149. favog says

    Okay, I looked at it again. ALL OF THE ANSWERS BEFORE THIS ARE WRONG!

    Yes, ten out of our hundred nights are break-ins nine of which are reported. Of the remaining 90 nights, 80% of them are false positives, so that’s 72 false positives. Add them to the 9 true positives, that’s 81. 9 true positives divided by 81 positive reports is … still 1 in 9. My earlier errors balanced out to give the correct answer.

  150. Narf says

    Why are you removing the 10 nights on which there was an actual theft attempt? You can still get a false positive on those nights, either before or after the actual theft attempt. Even removing the night that the car is actually stolen isn’t necessary, because you aren’t going to get any more alarms of either sort, after that.

    The lack of consideration of both a false alarm and a theft attempt on the same night is what makes the 11.11111% figure wrong. We’re talking about actual alarm events in the problem, not just the night that an alarm goes off, as a whole.

    You have to address the full 100 nights when figuring in the false alarms. That gives you 80 false alarms per 100 nights.

  151. Narf says

    @154 – fgt

    Oh and something that theists and apologists do that makes them come off like suspicious assholes to me, take their opinions and subjective ideas and present them as though they were empirical fact.

    That and endless hypocrisy. Going back to William Lane Craig, he starts out Reasonable Faith, complaining how hard it is to find debate opponents who comprehend basic logic, which he, as a professional philosopher, has a solid lock on.

    He then goes on to demonstrate, throughout the book, that he doesn’t comprehend the most basic logical fallacies. Almost the entire book is full of nonstop appeals to consequences. “If A, then B; and B makes me feel bad, so A can’t be true,” has never been a logically valid argument, but WLC seems to think it is.

    If you’d like another fantastic demonstration of apologist hypocrisy, Steve Shives did a fantastic examination of Carl Gallups’s book, The Magic Man in the Sky, in which Gallups attempts to tear down what he calls a straw-man that atheists construct, while violating the straw-man fallacy himself, worse than anyone I’ve ever seen. It starts getting particularly bad starting in chapter 4 or 5, I think, in the second video in that series.

  152. frankgturner says

    @ Narf #161
    As I said before, WLC is no professional philosopher, he’s a politician. Frankly his whole appeal to Divine Revelation is essentially saying that under no circumstances can he be wrong or factually incorrect even when presented with facts that contradict his position. In a nutshell that is the repeating a lie until people believe you concept (no need to Godwin, you know who said it) packaged with different words. If he really wanted to comprehend science he would contemplate falsifiability.
    .
    It is assholes like him that have often made me think that if there is justice in this universe then WLC needs to be publicly designated cruel and unusual punishment. He’s a bully though and it is what bullies do. If anyone drives me away from Xtianity and paints a picture of Xtianity as a cruel viscious cult it is WLC.

  153. Narf says

    Nah, you run into the same problem with cancer screenings. That’s one of the reasons that doctors tell you to remain calm, while they check again to make sure. The incidence of people with cancer is pretty low, below the age of 70 or so. The incidence of false positives is higher than the rate of people actually having cancer, I believe.

    Breathalyzers are even worse, since if you just brushed your teeth and used Listerine, you could blow above the legal limit, without even getting a measurable amount of alcohol into your system.

  154. Narf says

    @162 – fgt
    Yeah, I know. No one who’s read one of his books with the slightest shred of skepticism would mistake him for a real philosopher. He’s just one of the people that is nice to hold up, along with Ray Comfort, and now Carl Gallops, whom I’ve recently discovered.

    Forget any arguments against Christianity and religion in general. Those people. That’s why I’m an atheist.

  155. Monocle Smile says

    @Narf
    The Listerine thing was a problem in Bay City, Michigan for a bit. They instituted a city ordinance that allowed the police to subject anyone in a public area to a breathalyzer without probable cause, and if you were a minor and blew a 0.00, it was an automatic 30 days in custody.

    This didn’t last long, but it was rather frightening response to the abolition of the “body is a container” MIP law.

  156. Narf says

    Well, I’m sure they were only catching black and brown people in that net, so the over-the-top execution of that particular crackdown went according to plan.

  157. favog says

    Narf, I was wondering when you first mentioned Gallups if you had just listened to Steve Shives on the subject. I did, and … there are not words.

    9 nights of break in detected,+ 1 night of break in not detected, +72 nights of no break in but the detector goes off any way, +18 nights of no detector or break in = 100 nights. You have to multiply the 90 crime free nights by by 80% just like you multiplied the 10 crime nights by 90%. Answer is 11%.

  158. frankgturner says

    @ Narf #165
    Yeah WLC definitely takes advantage of the knowledge that the average person is not all that well educated, much less skeptical (well the average US American citizen, but that is changing). Despite his having a legitimate degree (unlike Hovind), WLC still is not versed in the sciences and demonstrates at best a high school level comprehension of a good portion of what he talks about, which is more than his constituency.
    .
    Metaphorically speaking WLC is the one eyed man among a group of blind people. Those of us with two eyes know that he is bullshitting his blind followers.
    .
    I’ve said the same thing many times, people like Comfort and Craig and Hovind don’t attract me to Xtianity, they drive me away from it by acting like incredible assholes.
    .
    @ Carol Spelling
    You might like my website. I just got started and am trying to make a bit of money from it if I can. I will be posting neat useful information as is characteristic of me looking things up on a daily basis (as I did growing up).
    Https:/learned.WordPress.com

  159. Narf says

    @168 – favog

    Narf, I was wondering when you first mentioned Gallups if you had just listened to Steve Shives on the subject. I did, and … there are not words.

    Yup. The defective thinking that that guy displays is just mind-numbing. The part about the Earth being designed for humans, later in the book, was one of the most stupid things I’ve ever heard.

    “If you removed humans from the ecology, the Earth would continue along just fine. If you removed all of the plants and animals from the Earth, humans would soon die. Therefore, clearly the Earth and the ecology were designed for humans.”

    How stupid/dishonest do you have to be to even come up with an argument like that, never mind presenting it in a book and expecting people to take it seriously?

    9 nights of break in detected,+ 1 night of break in not detected, +72 nights of no break in but the detector goes off any way, +18 nights of no detector or break in = 100 nights. You have to multiply the 90 crime free nights by by 80% just like you multiplied the 10 crime nights by 90%. Answer is 11%.

    No, it’s not. You’re reading the problem wrong.

    “You wake up one night and hear your alarm. What are the chances that a thief is breaking into your car ?”

    We’re looking at an instance of the alarm going off … unless you think a thief is going to try to steal a car which already has the alarm going off, because of a cat jumping on the hood.

    Alarm goes off … you wake up … when that alarm went off, was the thing that set it off a thief or a cat or truck or something? If you’re treating it as a nightly thing, that you would leave the alarm going all night, no matter what set it off, then you aren’t following the scenario laid out in the problem.

    You aren’t counting instances of a cat setting off the alarm on nights when a thief didn’t try to steal it. You’re counting nights when a cat set off the alarm, on both nights when a thief set off the alarm earlier (or later) and nights when a cat set off the alarm, when a thief hadn’t made an attempt. You can’t subtract the 10 nights that a thief tried to steal your car, when accounting for the cat. You can have both a false positive and a true positive or false negative in the same night, after you deactivate the alarm from the first instance of it going off.

    You might be getting hung up on the representation of the number of incidents per 100 night period. That’s just a representational artifice that represents frequency in a way that makes it easier for people to understand. If you remove the ten nights of actual theft attempts, you’re making it non-representative of the frequency.

    Treat the problem the other way around, and you’ll see the issue. Okay, so 80 nights out of 100, a cat sets off the alarm. On the remaining 20 nights, a thief tries to steal the car 2 of those nights, and the alarm goes off 1.8 of those times. 1.8 / 81.8 = 2.2004%. That’s … not right.

    This is a measure of frequency and probability, not of instances of occurrence over 100 nights. You have to count all 100 nights to discover the frequency of both possibilities, in order to see what the odds are of either possibility happening first.

  160. Narf says

    @169 – fgt

    Despite his having a legitimate degree (unlike Hovind), WLC still is not versed in the sciences and demonstrates at best a high school level comprehension of a good portion of what he talks about, which is more than his constituency.

    Yeah, I’ve seen a few physicists going over his grasp of cosmology. It’s pretty pathetic. The guy has no idea what he’s talking about, and it’s so freaking obvious, after listing to someone who does, for 5 minutes.

    It’s even more galling hearing WLC talk about how those who oppose the absolute reliability of the gospels don’t understand how historians evaluate history. Are you an historian, Dr. Craig? What is your degree in again? Dude, shut the fuck up and sit down, when you’re saying things so stupid that even a layman should know better.

  161. Narf says

    (continuation of #171)
    Look at that last paragraph in #171, favog. I think that highlights the issue with the way you’re doing things … or at least that there’s something very wrong there.

    In calculating frequency, we aren’t adding up to 100 days, with the two possible events. We’re figuring in the number of instances of the two possible events occurring within the same 100 day period, as a representation of frequency. There’s no subtraction.

  162. frankgturner says

    @ Narf #172
    TheYeah that just makes me laugh. The “absolute reliability of the Gospels”? What the hell is that supposed to mean?
    .
    There are even renowed theologians that don’t claim infallibility of the gospels, they contradict each other for crying out loud! Like I said, he’s a politician. Pay no attention to the man behind the curtain.
    .
    It is funny how the fundies give us their best and brightest and THIS is what they come up with. Of course, if it were not him some other dork would be running that dog and pony show.
    .
    I was listening to the non prophets the past couple of weeks and they talk about how atheism is a reminder to many Xtians of the grim finality of death (based on a recent published study). I think now it is more of a reminder how prices like WLC have no solid fact to base their ideas on so they have to deceive their constituents to even have a voice.

  163. Narf says

    Heh heh heh.  I know that, Frank.  Now see if you can explain it in such a way that the vapid Christians who venerate Craig will understand.

  164. frankgturner says

    Unfortunately Narf I think the vapid Xtians are like vapid conservatives who are like any other vapid person who has emotionally dedicated themselves to a cause. They believe that they are their ideas and that without those ideas they will simply vanish or become something they would hate or a miserable version of themselves (or all of the above). What they really need is an emotional turn around, something that they deeply care about and will loose if they push their agenda for which they cannot claim they are personally responsible.
    .
    I think that is what my cousin’s ex husband was going through when I tell that story about his son and the heart surgery. He wanted to scream that evolution was false and that they should not perform that surgery on his son but he had to listen for hours upon hours and demonstrate a working understanding of evolution given that his son would die without the technique based upon it. If you brought that topic up he would fall silent (sometimes for more than just that night according to my cousin). He had to be right and could not swallow his pride but it was as though if he did he would die as a person or his life would become meaningless. And if he could not shut up and refused to accept the procedure his son died, He could have done what a lot of creationists do and come up with some bullshit apologetics argument out of his ass, but I think the real possibility of loosing his son stood in the way of that.
    .
    As James Randi said, people like him “need” to believe. They have built their whole personality and every conception around an idea. Craig and his followers built their world around an idea and they just can’t seem to find the possibility that their lives could have meaning that goes beyond that.

  165. frankgturner says

    Oh and I don’t think they are vapid Xtians because they venerate Craig, I think they venerate Craig because they are vapid Xtians. Eliminate Craig and they will venerate someone else who feeds them bullshit.

  166. EnlightenmentLiberal says

    @frankgturner in 153
    Couple quick points.

    Factual matter first.

    You may not have intended to sound dogmatic EL, but I think you did despite my agreement with your general premise and mathematics.

    We don’t have to explicitly consider the likelihood of oddness when we are looking at the likelihood of even vs not even.

    We recognize that explicit evidence that could demonstrate the probability of a god being guilty or not guilty of existing, while it does influence the implicit probability of a god being innocent or not innocent of existing, we are making explicit statements and only making explicit considerations. Making explicit claims regarding the innocence or non innocence of a god existing is a separate consideration.

    You say that you agree with my positions, but the third quote above stands in stark opposition to the point I have been trying to make. There is a core philosophical difference, a core epistemic difference. It thus seems that you do not understand anything of the point I was trying to make. That makes me sad. At the very least, I have failed as a communicator.

    Moving to the tone discussion.

    I am not basing all of my posts to Matt off of content merely from this thread, or from the show of this thread which I have not watched. My posts to Matt are informed by a long history of Matt saying similar things. Perhaps because I (probably) have watched much more of Matt, both the AXP show and elsewhere, I have a better informed opinion, and perhaps for this reason you are more willing to grant the benefit of the doubt where I am not.

    You also seemingly disagree with several of my positions. Worse, you seemingly don’t understand several of my positions, probably because I have been a bad communicator. Apparently I have appeared to you to be someone who has hounded what appears to be irrelevant points because we are all in agreement (whereas we are in stark disagreement).

    For these reasons, I have come off to you as a bully and dogmatic. Let me put it plainly: I don’t care that I have come off as a bully and dogmatic if it is only because of these reasons. Sorry. Let me put it another way. You are saying that anyone who is confident in their assertions, and possibly who also disagrees with some consensus, is a bully and dogmatic. Again, to put it plainly: I strongly disagree.

    Now, I believe that there are other ways to be a bully and to be dogmatic. However, I don’t believe that I have been those things.

    Perhaps you felt bullied, and for that I offer my condolences, and I will examine my future actions to see if I can reasonably avoid making you feel bullied, but I’m not going to allow a heckler’s veto against what I consider to be eminently reasonable behavior.

    If anything, I need to apologize for being a bad communicator and letting things reach this point where we have talked past each othe so badly. However, at the moment, I don’t know how else I might explain myself.

    PS:

    Corwynn was not being a “bigger” asshole, he was being the same level of asshole to see if it would cast a reflection.
    I don’t think that you intended to come off as an asshole EL,

    Strongly disagreed. Intent is not magic, but intent matters. Had I intended to bully someone, it would be necessary to apologize.

  167. frankgturner says

    @ EL
    Perhaps the issue of communication comes from a lack of uniformity of word usage. So to some degree it is a matter of semantics. Of course as Matt says many times (something I had not heard before he said it but it makes so much sense) words do not have inherent or implicit meaning but have meaning by consensus. If one refuses to use words the way the majority of a particular group does then one is going to wind up a bad communicator. As an Aspie who went through what seems very similar to what you are going through (hence why I have thought you may be as well and have indicated so, several times), I recognize that sometimes one needs to succumb to the way a group uses certain words vs dogmatically insisting on my definitions being the “correct” meaning (something you personally have done several times in here).
    .
    (This is going to take several posts so bare with me as I post this and work on the next bit).

  168. EnlightenmentLiberal says

    Quoting me:

    Let me put it plainly: I don’t care that I have come off as a bully and dogmatic if it is only because of these reasons.

    Perhaps you felt bullied, and for that I offer my condolences, and I will examine my future actions to see if I can reasonably avoid making you feel bullied, but I’m not going to allow a heckler’s veto against what I consider to be eminently reasonable behavior.

    The first quote is a little over the top. Sorry. The second quote better represents my position.

  169. frankgturner says

    @EL
    All right here goes.
    For starters I have listened to a fair amount of matt and I have listed in other places outside of TAE. I have listened to his Youtube page and heard his talks on The Thinking Atheist and with Aronra, among others. I am going to go out on a limb here and hypothesize that maybe the reason I am more willing to give him the benefit of the doubt is that I do not perceive him saying or implying what you think he is implying. Or perhaps I do perceive it as a possibility whereas you see it as being stated definitely. That is why I tried to make a distinction between implicit vs explicit statements.
    .
    Let me try to get you to perceive where I am coming from and maybe you can give me an example to try to help me to perceive where you are coming from. I learned to read very early, almost as fast as I was learning to speak and walk. This is not uncommon for an Aspie (math came quick to me as well). I read books )simple books mind you) early on and came to think of word definitions in the way one might think of them in the dictionary (I often had one handy). I would notice people using words in speech in ways that did not agree with their textbook definition a lot. It took me time to realize that (as Matt has put it but I never had such a good way of describing it before watching him a couple of years ago) “Words have meaning by consensus.”
    .
    That was pretty powerful to me as I always thought that people were “supposed” to use words the way they were in a dictionary and not come up with their own usage. What I came to realize over time is that language is dynamic, changing and flexible. You have to change with it. I have a friend who adapts to new words and their usage socially very fast and will often be using expressions long before I am. Example: a couple of years ago I heard her using the word “sauce” after a lot of things, i.e.: “awesome sauce.” That seemed weird and kind of foolish to me and it took me a bit longer than her to get used to that. She is generally more dynamic about language than I am.
    .
    I had to learn that there were people like this and that I was NOT one of them. My mind is not as dynamic and flexible as theirs are, but in many ways I try to be. I had to learn to be opened to different models of how to think of things and how words might be used (among other things).
    .
    So perhaps you and I have different ideas about what constitutes explicit vs implicit statements. Perhaps everyone here has a different idea of what is meant by those terms. I don’t know as I don’t speak for everyone, but I DO recognize that what influences other people’s understanding of words and ideas is THEIR background which is going to be different from mine. We may have to unpack “implicit” vs. “explicit” a bit and recognize that no one is right or wrong. There is not some worldwide organization that polices communication and assures that all people are using the words in the English language the same way so we are going to need to open our minds with each other, I often couch my words with a lot of “maybes” and “perhaps” because I recognize that while I might think someone is wrong, it might just be my perception based on my understanding.
    .
    Part of what I do is continuously think and re-think my models. I am constantly opened to new ideas and evidence that may demonstrate that even the based assumptions that I am making (the meanings of words and how they are used as an example) may be faulty. I said on another message board here about how talked to a theist some years ago who said to me. “well if god did not create the universe than who did?” To which my response was one I had heard Matt give (albeit I had thought of this a few times before), “does it have to be a ‘WHO’?” The theist had never realized that he had always accepted the idea of something greater than us having a sentient mind and had never even considered the possibility that it did not. Though not against evolution he wondered how life began and I said, “abiogenesis.” He said, “no … evolution, what is abiogenesis?” I had to talk to him a bit to determine that he had early cosmic inflation, abiogenesis, and evolution all confused into one idea, not realizing that they were separate principles. I sat down with him and at least was able to put ideas into words well enough to get him to comprehend that they were different principles, without even stating whether they were right or wrong. It got him to re-think his models for how he viewed things.
    .
    That would not have worked very well as far as communication goes if he were unwilling to re-think his models. If he had insisted that he was right and that based on what people had told him (his ministers I presume). He was able to put into words what he believed (which is how I figured out that he had 3 different scientific principles confused into one, which sounds like typical apologist stuff based on my personal experience). I may not have your idea clearly, but it seems like no one else does either. Given that no one seems to comprehend what you are talking about, frankly it seemed dogmatic to insist that you were right about Matt being wrong when you yourself can’t even clearly put into words why or how he is wrong in a way that anyone understands but you. That seems bully some and while I don’t speak for everyone else, it sounds like i am not the only one who got that impression.
    .
    I AM willing to entertain that I don’t comprehend you, but part of the way that I am TRYING to comprehend you is to put myself in your shoes and try to think of the conditions under which you might be correct. That is what i was trying to do in # 153 and I obviously did not succeed. In other words I did not dogmatically insist that I was right and that you were wrong if you disagreed with me, I hypothesized that YOU were right and tried to consider the conditions that made you right. I did not know if I got it correct which is why I couch the terms a lot. You might, just might, do better at your communication by extending the same courtesy. Think of why I might be right, think of why corwyn or Matt might be right. Get inside their head and think of the conditions, WHY are they thinking what they are thinking?
    .
    Unpack your statements a bit and try to prove yourself wrong. You gain no understanding of how better to communicate by beginning with the conclusion that your are right and that if others can’t understand that it is their fault. That is what apologists and theists do like WLC and THAT is what you were sounding like to me and I would not be surprised if others on here feel the same way. Open your mind to the possibility that you might be wrong and it MIGHT help you better figure out why you are right, or you might even see that you have not been.
    .
    I will try to put it into Bayesian Terms. Let us say that someone provides a statement to me regarding evidence for god. I.e.: they give me evidence that a sentient mind can exist without a physical body. Based on my personal understanding of god and that of many others, THIS is a necessary construct in the logic for a god. Now this does not prove that one exists, only that one is POSSIBLE. I don’t know how this effects the chances that a god exists are true in the Boolean sense, but I know that if X = god exists and A = a sentient mind is capable without a physical body and I have some degree of certainty of the chances of A, i.e. A is possible, and A is necessary for X, then X is possible.
    .
    The thing is, when I compare it to a court case, I am NOT saying that X = god exists. I AM saying (If I get Matt correctly) that X = the possibility that a god is guilty of existing. I am NOT making a DIRECT statement about Y which will be Y = the possibility that god is innocent of existing. I DO recognize that the chances of X and not X, i.e.: that chance that god is guilty of existing = X and the chance that god is NOT guilty of existing = ~X do influence Y. Because I and NOT making a direct and intentional statement about Y I consider the influences of Y and ~Y the chances of X and ~X to be implicit and NOT explicit. I will consider Y and ~X in my calculations of X and ~X when examining that principle which I am focusing upon.
    .
    If I decide to examine Y and ~Y THEN and only then will I consider myself to be explicitly, that is directly and intentionally and I am FOCUSED upon Y and ~Y, to be examining Y and ~Y explicitly, that is the chances of god being innocent of existing and the chances of god being NOT innocent of existing. X and ~X have strong implicit (indirect, but nonetheless relevant) influences upo the equation to determine Y and ~Y.
    .
    Now you may have presuppositions for gods existence vs non existence similar to that of the gum balls, With such a tight system if the gum balls where X = the chances that the gum balls are even and Y = the chances that the gum balls are odd, X = ~Y and Y = ~X. I don;t feel that the god question is that tight without a LOT of unfounded presuppositions. Maybe you do and you are entitled to that. I don;t think of that as BAD thinking though as it acknowledges another way of looking at things.
    .
    Tat is actually the impression that I am getting from you, somehow what you think of Matt is that on some level he is speaking in absolute terms. And you insist this and do;t ask him based on an impression that you get from several shows? And you don’t see that as dogmatic to assume that the person you are accusing means something that you disagree with when you can just ASK them if they mean it directly?
    .
    Maybe it is because you can’t figure out how to put it into words that make sense to anyone but yourself. Maybe that means that you need to re-examine if it really makes sense to you on a logical level or if it is just a feeling that you get. Although I don’t consider this direct evidence, when you said, The problem with the gumball example is that it’s hard to express this to someone who does not already understand . Most of the time since I have been reading in the past couple of years, when i hear the phrase “hard to explain to someone who does not already understand,” it sends red flags up and in many cases the person seems to have no idea what they are talking about.
    .
    Also, in response to You are saying that anyone who is confident in their assertions, and possibly who also disagrees with some consensus, is a bully and dogmatic no that is not what I am saying. However, IF a person is ABSOLUTELY and TOTALLY confident in their assertions and completely and totally unwilling to entertain even the slightest degree of possibility, in their own mind or others, that they might, even to the smallest of degrees, be incorrect or just flat out wrong, well that makes you just like the theists and apologists. They insist that their belief in their god MUST be RIGHT and that anyone who disagrees is wrong or “just does not understand why.”
    .
    When you said, I still believe I am right on all of my points in the thread (give or take a very minor mistake I’ve overlooked)
    Did it even occur to you in the slightest possible way that you MIGHT be wrong? Maybe you are not, but having the humility, heck the open mindedness to consider it as a POSSIBILITY, however minute or large it may be is a BIG part of what goes on in here. You obviously must be wrong if not in fact than in your way of communicating it to others. If you are not wrong and you want to communicate better to others you might try to consider, SERIOUSLY consider the possibility of being wrong, perhaps even drastically.
    .
    I want to bring up a point that Matt has made in several of his talks about debating. You say that you have listened a lot and maybe you remember having heard this, maybe not, If you have had such drastic experience listening to him and don’t remember this then I suggest listening to him more as I have heard him say this more than once and it is pretty important. He said (paraphrasing) that one of the biggest parts of debating is knowing your opponents position. He said that as a good debater you need to know your opponent’s argument so well that you could practically make their argument for them. He did that in front of Sye Ten Bruggencate. He knew StB’s position so well that one barely needed to listen to what StB was saying. If you listened to this episode the same thing was done with WLC.
    .
    So if you are so confident that your view is correct EL why not make an effort to argue the OTHER side and see if you can make sense out of it? If you can’t, well maybe you don’t understand it well enough, which might be why no one seems to agree with or even comprehend your argument against it. I a really amazed that you never even considered for a moment that maybe your disagreement with Matt comes from a place of misunderstanding. You might really have to look deep inside at the base model though and unpack it.

  170. EnlightenmentLiberal says

    @frank

    I’m trying to make this short. I don’t mean this as an insult or anything, but I don’t want to engage in a pissing match. It doesn’t matter who is right about what Matt actually thinks and what is the best or reasonable interpretation of what he has said on the matter. At least, let’s move on to our disagreement.

    What you wrote is rather long. I considered it at length, and wrote an even longer reply. I cut that back for both of our interests. Let me know if you think I didn’t answer something important. Sorry.

    Maybe you are not, but having the humility, heck the open mindedness to consider it as a POSSIBILITY, however minute or large it may be is a BIG part of what goes on in here.

    “Possible thus probable” fallacy. It is true that I may be wrong, but that alone is not enough to show that I am probably wrong, or that the chance that I am wrong is in any way significantly likely. For comparison, I also might be wrong about the age of the Earth, but it does not follow that the chance that I am wrong about the age of the Earth is in any way significantly likely, and it’s not a reason for me to be meek about asserting the age of the Earth.

    Again, I have been asserting but two principles.

    (∀X is a proposition)(P(X) + P(¬X) = 1)

    In English: For every proposition, a rational actor’s estimation of the likelihood that X is true plus the estimation of the likelihod that X is false must equal 1.

    (E is evidence for proposition X) ↔ (∀Y is an alternative proposition)(P(E | X) > P(E | Y))

    In English: Something counts as evidence in favor of a proposition if and only if it is expected on the proposition and not expected on any alternative proposition. More generally, something counts as evidence in favor of a proposition if and only if the degree of the expectedness of the evidence if the proposition is true is greater than the degree of the expectedness of the evidence on all alternative propositions.

    Are you denying either of these principles?

    If you agree with these principles, then it seemingly follows that in order to consider any claim, it is requisite to concurrently consider all alternative claims. Do you agree here?

    Previously, you said this:
    bolding added

    We recognize that explicit evidence that could demonstrate the probability of a god being guilty or not guilty of existing, while it does influence the implicit probability of a god being innocent or not innocent of existing, we are making explicit statements and only making explicit considerations. Making explicit claims regarding the innocence or non innocence of a god existing is a separate consideration.

    Given what I’ve said, what you said in bold is wrong. Everything after that is a failure to communicate on one or both of our parts. I suspect in large part me.

    PS: I do not understand what “implicit vs explicit” has to do with anything. I don’t care if you talk aloud about your reasoning, or if your reasoning is internal. I don’t care if you formally do Bayes equation every time, or do approximations and heuristics. I care if you blindly accept evidence for a proposition without considering at some level, explicit or implicit, if that evidence is a equal or better fit for alternative explanations. I care if you somehow hold estimations, probabilities, odds, of the truth and falseness of a claim which do not add up to 100%.

  171. EnlightenmentLiberal says

    Addendum. I meant to emphasize the following words:

    Quoting me, bolding added:

    If you agree with these principles, then it seemingly follows that in order to consider any claim, it is requisite to concurrently consider all alternative claims. Do you agree here?

    Quoting you, bolding added:

    Making explicit claims regarding the innocence or non innocence of a god existing is a separate consideration.

    It is not a separate consideration. It is an intimately intertwined consideration. It’s a logically inseparable consideration. For example, I don’t need to consider the number of planets in the solar system to build a car. Those are separate considerations. When considering whether a god exists, “no god exists” is the same consideration (with a logical negation). For more complex alternative hypothesis spaces, each space of competing hypotheses are not separate considerations but different facets of the same consideration. This is embodied in Bayes equation.

  172. frankgturner says

    @ EL
    I think that we are getting somewhere with this and I think I see where you might be misunderstanding.
    .
    The courtroom analogy is proposed because we are NOT a war of the embodiment of all things that make up a proposition. While we know that P(X) + P(~X) = 1, we don’t know P(X) or P(~X). The proposition of Innocence is NOT considered to be The diametrical opposite of P(X) such that P(Y) = P(~X). So P(X) +P(Y) /= 1.
    .
    The court is not asking you to consider ALL alternative propositions, merely the ones that we know and can be or are resonobly presented at this current time. Part of scientific endeavors is recognizing that we accept a hypothesis as a theory when it has good evidence to support it. We are meant to keep our minds opened to all evidence, including evidence that may support opposing principles that we are not currently aware of.
    .
    So yes it is requisite to consider ALL alternative claims, if we know what all of those alternative claims are . If we DON’T know what all of those alternative claims are, we STILL need to make estimates based on what information that we DO have. We do so until other information comes along. Many scientific endeavors are about continously looking for that information because it is recognized that we DON’T have all of the information to determine the likelihood of all alternatives.

  173. frankgturner says

    P.S.: This is why I think that the gumball analogy is a bad one. Given the assumptions made the Probability of even (X) is the diametrical opposite of the Probability of odd (Y) such that P(X) + P(Y) does = P(X) + P(~X) = P(Y) + P(~Y) = 1.
    However, in a courtroom it DOES matter whether you make those calculations internally or externally and the probabilities of guilt and non guilt vs the probabilities of Innocence and non innoncence don’t add up so nicely. The court is mainly interested in your estimations for guilt and non guilt given the information that is presented.

  174. EnlightenmentLiberal says

    In reverse order.

    So yes it is requisite to consider ALL alternative claims, if we know what all of those alternative claims are . If we DON’T know what all of those alternative claims are, we STILL need to make estimates based on what information that we DO have. We do so until other information comes along. Many scientific endeavors are about continously looking for that information because it is recognized that we DON’T have all of the information to determine the likelihood of all alternatives.

    Agreed.

    The court is not asking you to consider ALL alternative propositions, merely the ones that we know and can be or are resonobly presented at this current time. Part of scientific endeavors is recognizing that we accept a hypothesis as a theory when it has good evidence to support it. We are meant to keep our minds opened to all evidence, including evidence that may support opposing principles that we are not currently aware of.

    Agreed … I think.

    I think that we are getting somewhere with this and I think I see where you might be misunderstanding.
    .
    The courtroom analogy is proposed because we are NOT a war of the embodiment of all things that make up a proposition. While we know that P(X) + P(~X) = 1, we don’t know P(X) or P(~X). The proposition of Innocence is NOT considered to be The diametrical opposite of P(X) such that P(Y) = P(~X). So P(X) +P(Y) /= 1.

    And you lost me.

    The job of a juror in court is to create a (justified) estimation of the epistemic probability of guilt, which is the same thing as the epistemic confidence of innocence under statistical complement, e.g. P(innocent) = 100% – p(guilty). That’s the job of a juror. If they have a particular estimation of the likelihood of guilt which is greater than the “no reasonable doubts” threshold – whatever that number is, maybe 99%, I don’t know – then they should vote “guilty”. Otherwise they should vote “not guilty”.

    In order to create a (justified) estimation of the epistemic probability of guilt, they need to try to create an exhaustive list of scenarios which are concordant with the evidence. Then they need to churn out probabilities P(evidence | scenario) for every scenario. Then they need to run the numbers via Bayes equation to calculate P(scenario) for every scenario. Then just add up the probabilities of the scenarios to determine epistemic probability of guilt – again which is the same thing as the epistemic probability of innocence under statistical complement e.g. 100% – X.

    So, what do you mean by P(X)? I agree that estimating P(evidence | scenario) is the hard part where we need to apply intuition, philosophy, common sense, and other unscientific tools, but this unscientific aspect of estimating probability is there in every scientific endeavor. Formal bayesian reasoning just makes that step explicit.

    Did you mean P(scenario) ? Once you have estimates of P(evidence | scenario) for the scenarios, then you just run the numbers via Bayes equation to calculate P(scenario).

    While I do not suggest formally using Bayes equation for everyday events, if I was ever on a jury I probably would formally use Bayes equation and put actual numbers to particular pieces of evidence on particular scenarios. (I’m sure my fellow jurors would hate me, likely being innumerate and all.)

    PS: If I was in a situation where I felt uncomfortable about claiming that I had a sufficient proper exhaustive list of scenarios, I might create another scenario called “unknown”. That would be fun.

    PS: I’d probably study the method and objections a bit more before I actually used it in practice on a jury. I am partially swinging from the hip here.

  175. frankgturner says

    @ EL
    And that’s where you misunderstand how a courtroom works. In a courtroom, P(innocence) /= 100% – P(guilt). P(not guilt) does = 100% – P(guilt) but P(innocence) /= P(not guilt).
    .
    That is why the court does not declare formal innocence, only guilt and non guilt whose calculations could be wrong given a recognition of limited information.

    We only make estimates of P(guilt) and P(not guilt) because we recognize that we have limited information to make those calculations and we recognize that we may not have all information.
    .
    If you want to insist that P(innocence) = P(not guilt) then go ahead, but that will not help you to communicate with those of us who recognize that this is not the case as far as the court is concerned. Perhaps it would you you well to consider why this is the case. As far as the court is concerned innocence is NOT the statistical compliment of guilt. However, the Probability of guilt is influenced by the Probability of Innocence even if they are not compliments (again as far as the court is concerned).

  176. EnlightenmentLiberal says

    @frank
    You still have a particular stumbling block. Let me try to explain it again.

    In reality, either the person committed the crime, or they didn’t. There is no in-between. (Let’s ignore corner cases for now.)

    For example, for a simple murder via a knife with multiple stab wounds, by far the only plausible scenario is that someone stabbed the victim. Right off the bat we can rule out any other scenario barring extreme evidence to the contrary. The question then becomes “who did it?”.

    Again for example, let’s assume the evidence is conclusive that it was either Alice, Bob, or Carol. Again, the facts of the matter in this hypothetical is that either Alice did it, or Bob did it, or Carol did it. There is no other option (under the hypothetical evidence).

    P(Alice is guilty) + P(Bob is guilty) + P(Carol is guilty) = 1.

    We can write some other equations too.

    P(Alice is innocent)
    = 100% – P(Alice is guilty)
    = 100% – (100% – P(Bob is guilty) – P(Carol is guilty)
    = P(Bob is guilty) + P(Carol is guilty)
    = P(Bob is guilty or Carol is guilty) … [Assuming independent events, e.g. assuming only one murderer]

    The above equation also makes sense. The probably that Alice is innocent should be identical to the probability that Bob did it or Carol did it.

    Your problem is that you are thinking in terms of epistemic confidence of physical probability. Knock that shit right off. That’s the wrong way to think about it. You’re not estimating your confidence of a particular probability that Alice is guilty. No. You are directly giving your gambler odds for the actual truth that Alice is guilty. Suppose you’re in a (weird) game show where there is irrefutable proof, but that is withheld from you, and you only get pieces of the evidence. If they offered you a bet about the guilt of Alice, you would have to determine an estimation of the likelihood of guilt, a single number, in order to inform your bet. It’s the difference between accepting a $1 to $1 better, vs a $1 to $10 bet. That’s what we call “odds”. If you think there’s a 50% chance that Alice is guilty, then taking a bet of “you pay $10 on a loss, and you gain $1 on a win” is a really bad bet. However, If you were 99% confident that Alice is guilty, then that’s a great bet.

    P.S.: This is why I think that the gumball analogy is a bad one. Given the assumptions made the Probability of even (X) is the diametrical opposite of the Probability of odd (Y) such that P(X) + P(Y) does = P(X) + P(~X) = P(Y) + P(~Y) = 1.

    This right here above. I’m pretty sure that’s your problem. Stop thinking like a frequentist. Stop thinking about physical probabilities. Stop thinking about physical laws as being absolute. You’re not providing a degree of confidence for the odds of guilt. You’re simply providing a degree of confidence of guilt. Full stop.

    The following are always true (under the assumption that exactly one of them is guilty):
    P(Alice is guilty) = 100% – P(Bob is guilty) – P(Carol is guilty)
    P(Alice is guilty) = P(Bob is innocent) + P(Carol is innocent)
    And so on. These are simply true statements about the proper gambler odds of uncertain facts about the real world. This is the information you would use if you wanted to be a rational gambler, and we’re all rational gamblers. That’s what a rational actor is – an agent who estimates odds and makes cost-benefit analysis based on their estimated odds.

    Suppose we found a murder weapon found in Alice’s home, That means P(evidence | Alice is guilty) is very high. Of course, that on its own is useless. We also need to argue that P(evidence | Bob is guilty) is a lower number. For example, perhaps we know that Alice and Bob live in separate houses. That is enough to conclude P(evidence | Alice is guilty) > P(evidence | Bob is guilty), and on that evidence alone, one should use Bayes equation to conclude P(Alice is guilty) > P(Bob is guilty).

    At no point am I considering anything like P(it is possible that Alice is guilty). That simply does not come up in the analysis.

    For example, if I had evidence that Alice was across the world at the time of the murder, such as live tv video of the person on the other side of the world, I wouldn’t muck about with P(it is possible that Alice is guilty) or P(it is not possible that Alice is guilty). Rather, I would use that to argue that P(evidence | Alice is guilty) is staggeringly small, which in turn will cause P(Alice is guilty) to be staggeringly small (on that evidence alone).

    To continue, video tape of Alice on the other side of the world drastically affects P(evidence | Alice is guilty), but it does not affect P(evidence | Bob is guilty) at all. However, the net result of Bayes equation is that P(Alice is guilty) will go down drastically, and P(Bob is guilty) will go up drastically. On just that tv show evidence alone, after running the numbers, I would conclude that P(Alice is guilty) is near 0%, P(Bob is guilty) is near 50%, and P(Carol is guilty) is near 50%. (This assumes fair priors.)

  177. EnlightenmentLiberal says

    @frank

    P(innocence) /= 100% – P(guilt). P(not guilt) does = 100% – P(guilt) but P(innocence) /= P(not guilt).

    I have no idea what you’re talking about. Either the person did the crime, or they didn’t. There’s no in between. Our estimations may be in between 0% and 100%, but the facts of the matter are that they did it or they didn’t.

    I simply have no idea what you are talking about. You’re having a different conversation than the one I’m having, and I don’t know what conversation you are having.

    P(innocent) is the estimation of the likelihood of innocent that a rational actor should hold based on the available evidence. P(guilty) is the estimation of the likelihood of guilt that a rational actor should hold based on the available evidence. It’s irrational to believe that it’s 40% likely that the person is innocent but 80% likely that the person is guilty. That makes no sense. Put another way, it would be irrational for a person to take a 1 to 2 odds bet that the person is innocent and also take a 1 to 2 odds that the person is guilty. Again, please think about this like a gambler and gambling odds.

  178. EnlightenmentLiberal says

    tl;dr
    If you are a betting person, you need to calculate the worst payout bet that you would be willing to accept when someone offers you a bet, and that depends on your estimation of the likelihood that you will win the bet. Again, suppose someone claims incontrovertible proof of innocence and guilt, but offers you a bet of $X net payout for innocence, and $Y net pay on guilty. In order to determine if you want to take that bet, you need to estimate the probability that you will that bet. You will arrive at a certain minimum betting odds that will be willing to take, for example 1 to 2 odds.

    Suppose the person offers you a another bet, a reverse bet, where they offer to pay you some money on guilty, and you win some money on innocence. Your earlier calculation will necessarily constraint what bets you would be willing to take.

    To put it one way, Bayesian reasoning is all about giving betting odds for the truth of propositions. And if I’ve done my fourth grade math right, that means if you initially were willing to go no worse than X to Y odds, it means that you should be willing to go no worse than Y to X odds for the reverse bet.

    PS: The fun part of a jury trial is that they do not return a number of confidence that the person is guilty. Instead, they return a single boolean, “confident that the person is guilty beyond all reasonable doubt”. They must map the confidence space of [0%, 100%] to [false, true]. One has to decide what “beyond all reasonable doubt” means. Perhaps it’s 99%. Thus, they calculate the confidence of guilt, and they return “guilty, not guilty” according to the following piecewise function:

    0% to 99% confidence of guilt –> not guilty verdict
    99% to 100% confidence of guilt –> guilty verdict

  179. EnlightenmentLiberal says

    Clarification / correction.
    P(Alice is guilty) + P(Bob is guilty) + P(Carol is guilty) = 1.
    If I remember my basic statistics, that equation is true iff there is exactly one guilty person. I mentioned this assumption at one spot, but let me make it clear. If two people conspired together, then the probabilities are not independent, and the equation gets messier. It’s stats 101.

  180. frankgturner says

    I am aware of stats 101. I would counter with stop thinking like an absolutist. You mentioned that intent is important. It is in a courtroom as well.
    .
    Yes there most certainly are situations where there is in between when it comes to whether a crime was committed. Just because a person has committed an act, for which there maybe Boolean confidence (they either did or did not do it) there may be variable probabilities as to whether that act is a crime. I am amazed that you have not thought about this or even read about it.
    .
    No it is not irrational to believe that there is 40% confidence that someone is innocent and 70 % that they are guilty. They are not compliments as far as the court is concerned and I am amazed that you have not figured out why.

  181. John Iacoletti says

    The point of the gumball analogy has nothing to do with probability. In the absence of any evidence, it is not rational to believe either of the following claims, even though we know that one of them must be true.

    – The number of gumballs in the jar is even
    – The number of gumballs in the jar is odd

    Without any evidence, you start with the null hypothesis.

  182. John Iacoletti says

    In the courtroom analogy, it’s not a confidence level of the person’s guilt, it’s a confidence level that the prosecution has proven the person’s guilt. There’s a difference.

  183. Hippycow says

    @John Iacoletti #194:
    The gumball analogy is technically correct, but it seems many people are thrown off by the fact that the odds in that case are 50/50. It might be better to use some case where the odds are more unknown.

    By the way, this important fact* should also be noted:

    There are exactly 15,747,724,136,275,002,577,605,653,961,181,555,468,044,717,914,527,116,709,366,231,425,076,185,631,031,296 protons in the universe.
    Not 15,747,724,136,275,002,577,605,653,961,181,555,468,044,717,914,527,116,709,366,231,425,076,185,631,031,297 protons. Not 15,747,724,136,275,002,577,605,653,961,181,555,468,044,717,914,527,116,709,366,231,425,076,185,631,031,298 protons.

    * “fact” should be used in the way The Professional Philosopher, William Lane Craig, PhD uses the word, of course. In that sense, we can take “fact” to mean precisely “not a fact.”

  184. frankgturner says

    @ John #194
    Actually without evidence, one of those two could be wrong unless we make the assumption that the number is non zero.
    .
    You make a nice point about the courtroom analogy and that does fit that it is more conditional than simply a probability of guilt. What I was getting at though is that innocence is not the compliment of guilt. It is not bad thinking to do a bit of research and find out why this is so. Only non guilt is the compliment of guilt and there are in between situations. It is not simply a matter of a crime was committed or not committed, it goes deeper than that and what I am getting at IS something taken into consideration in courts.

  185. EnlightenmentLiberal says

    @Frank

    Yes there most certainly are situations where there is in between when it comes to whether a crime was committed. Just because a person has committed an act, for which there maybe Boolean confidence (they either did or did not do it) there may be variable probabilities as to whether that act is a crime. I am amazed that you have not thought about this or even read about it.

    Again, it might help if you read what I wrote a little better. I have mentioned this several times, and at least once in the last few posts.

    For example, quoting me:

    In reality, either the person committed the crime, or they didn’t. There is no in-between. (Let’s ignore corner cases for now.)

    Also, There is a possible confusion on your part in terms. Let me repeat myself again. There is the state of affairs “not guilty”, and there is the verdict “not guilty”.

    As a working assumption for the sake of argument, either someone committed the crime or they didn’t. They are either innocent or guilty. They are either guilty or not guilty. They are either innocent or not innocent. This is simply a description of the event space. The set of all events which are not the guilty event equals the set of the innocent event.
    event space = { “guilty event”, “innocent event” }
    { X in event space : X != “guilty event” } = { “innocent event” }

    The “not guilty” verdict is not a description of the event space. The “not guilty” verdict is a description of the jury’s confidence about the truth of propositions “X is guilty”, “X is innocent”, “X is not guilty” and “x is not innocent”. In particular, a “not guilty” verdict implies the following Bayesian confidence levels (assuming 99% cutoff for “beyond reasonable doubt”):
    P(guilty) in the range [0%, 99%]
    P(not guilty, the event, not a the verdict) in the range [1%, 100%]
    P(innocent) in the range [1%, 100%]
    P(not innocent, the event) in the range [0%, 99]

    In context, P(not guilty the verdict) makes no sense. The thing that goes in the P() is an event in the event space. “Not guilty the verdict” is a description of confidence levels, not an event in the event space. You’re confusing descriptions of confidence with events in the event space.

    PS: If you were an outside observer, then “jury votes guily verdict” and “jury votes not guilty verdict” are events in an event space, a different event space, but that’s a different conversation from what we’re having.

    Also

    And that’s where you misunderstand how a courtroom works. In a courtroom, P(innocence) /= 100% – P(guilt). P(not guilt) does = 100% – P(guilt) but P(innocence) /= P(not guilt).

    No it is not irrational to believe that there is 40% confidence that someone is innocent and 70 % that they are guilty. They are not compliments as far as the court is concerned and I am amazed that you have not figured out why.

    Frank, you seem to be speaking proper English, but I do not understand at all what you are saying. Your positions in the quotes are contradictory. If your Bayesian confidences P(innocent) + P(guilt) are not required to sum to 1, then you are allowed silly Bayesian confidences like P(innocent) = 40% and simultaneously P(guilt) = 70%. You cannot have it both ways.

    You are making a claim that he is teaching people to think badly. Really? What evidence to you have that supports this claim?

    Case in point. Here is some “evidence”. The immediately previous quotes of you taken together are nonsense. I want to hope that it’s because of your confusion of “not guilty” as an event in the event space and “not guilty” as a verdict, but otherwise I am completely lost.

    Actually without evidence, one of those two [even number and odd number] could be wrong unless we make the assumption that the number is non zero.

    Protip: “0” is an even number.


    @John

    I’m not following you here at all. If you’re a prosecutor and you claim the defendant is guilty, and then fail in your burden of proof to show that he is guilty, you have done nothing whatsoever to demonstrate that he is innocent with any probability or confidence.

    I’m sorry. I don’t know how to respond. I never made that claim.

    In this context, I would define “demonstration” as a presentation of evidence and argument which changes someone estimation of the likelihood of the truth of some claim from near 50% to near 100%. (By the math, this also includes presentations of evidence and argument which changes someone’s estimations of the falsity of some claim, or changing someone’s estimations of a claim from 50% to something near 0%.)

    For a hypothetical prosecutor in a case who gets up and presents no evidence, I never claimed that this is a demonstration of innocence. It doesn’t match definition of “demonstration”. He didn’t present any evidence. He didn’t demonstrate anything.

    “Extraordinary claims require extraordinary evidence.” That’s simply a restatement or a way to capture part of the Bayesian logic that you have prior estimations (priors) about certain things. Certain things you have previously determined to be highly unlikely, i.e. extraordinary, and that means you have low priors for these events.

    We always have prior estimations of an event. In the extreme case where we know absolutely nothing, 50 50 odds are generally prudent.

    This hypothetical prosecutor didn’t demonstrate anything, which means that the prior estimations of guilt and innocent of the jury remain unchanged. Assuming a competent jury which didn’t start with priors of P(guilty) &gt 99%, that means that the jury remains unconvinced that the person is guilty beyond reasonable doubt. They retain their prior estimations of guilt and innocence.

    In the courtroom analogy, it’s not a confidence level of the person’s guilt, it’s a confidence level that the prosecution has proven the person’s guilt. There’s a difference.

    Wait what? If you are an ideal juror who does not bring in outside specific evidence and who follows the presented evidence to the logical conclusions, then those two things are the same number. I suppose there might be a difference if a juror will blindly consider only the literal exact arguments and scenarios as presented by the prosecutor (and defense?), but IMHO that’s a bad juror. I hope that you are going to go above and beyond the literal exact arguments and scenarios as presented by the prosecutor and defense. (Not evidence, but interpretations of evidence.)

    Are you really saying that if you were on a jury, and the prosecutor missed one minor, trivial detail of their argument, then you would ignore all of the evidence and vote “not guilty” ? Really? I think I am appalled.

    What do you mean with your above quote? What is this difference? Can you describe a specific example please?

  186. EnlightenmentLiberal says

    @John

    The point of the gumball analogy has nothing to do with probability. In the absence of any evidence, it is not rational to believe either of the following claims, even though we know that one of them must be true.

    – The number of gumballs in the jar is even
    – The number of gumballs in the jar is odd

    Without any evidence, you start with the null hypothesis.

    Again, terminology problem, and a problem of the way that you are thinking. Again, in Bayesian terminology, “confidence” and “probability” are largely interchangeable. Again, start thinking like a gambler. You already are a gambler. Everything you believe and know are not guaranteed to be certainly true. Instead, there are odds, or chances, or probabilities, that you are wrong. You need to get estimates for the odds that you are wrong in order to make cost-benefit analyses in order to live your life. Every single decision you make is made on incomplete information, and thus it’s analogous to a bet.

    If you had no evidence about the number of gumballs in the jar (besides an assurance that it’s a whole number and not a fraction), then you need better than 50 50 payout before you take a bet that the number is odd, and you need better than 50 50 payout before you take a bet that the number is even. 50 50 odds is equal to 50 / (50 + 50) = 50% probability which is 50% confidence.

    Suppose someone was playing a different game, where they were offering bets for “number is divisible by 3” and “number is not divisible by 3”, again with an assurance that the number is a whole number. Here, as a gambler, you need better than 2 to 1 payout to take the bet that the number is divisible by 3, which is 2 / (2 + 1) = 2 / 3 =approx 66% probability = 66% confidence.

    In this “divisible by 3” game, with no further information, I would have a 33% (approx) confidence that the number is divisible by 3, and a 66% (approx) confidence that the number is not divisible by 3.

    Again, predicted on the assurance that the number is a whole number. Even that is not held to an absolute degree of 100% confidence, but I’m approximating it as 100% for the purposes of presentation. If we wanted to be more precise, we could say that being informed and promised that the number is a whole number may be enough to warrant a 99.99% prior that the number is a whole number, which means my estimation “the number is divisible by 3” is (99.99%)(1/3), which is a number quite close to 1/3 e.g. 33% (approx).

  187. EnlightenmentLiberal says

    Suppose someone was playing a different game, where they were offering bets for “number is divisible by 3″ and “number is not divisible by 3″, again with an assurance that the number is a whole number. Here, as a gambler, you need better than 2 to 1 payout to take the bet that the number is divisible by 3, which is 2 / (2 + 1) = 2 / 3 =approx 66% probability = 66% confidence.

    That makes no sense. Let me try again.

    The proper estimations are:
    P(number is divisible by 3) = 1/3
    P(number is not divisible by 3) = 2/3

    The general rule of a rational bet is one where the payout is on average positive (or at least non-negative). In the terminology of gambling, a bet with payouts of 2 to 1 means that I will pay 2 dollars if I lose the bet, and I will win 1 dollar if I gain the bet, proportionally speaking. We can calculate my expected winnings of taking a 2 to 1 payout bet on the bet that the number is divisible by 3.
    E(winnings)
    = ( – 2 ) ( P (number is divisible by 3) ) + ( + 1)( P (number is not divisible by 3) )
    = ( – 2 ) (1/3) + ( + 1)(2/3)
    = 0

    That’s a 0 net expected payout, which means that’s the minimum threshold payout ratio for taking the bet.

    Note the relation. I have the estimation P(number is odd) = 1/3, which is the same thing as saying that I estimate it as 1 to 2 odds that the number is divisible by 3. “1 to 2 odds” means 1 / (1 + 2) = 1/3 probability. The break-even payout ratio is 2 to 1 odds. That’s not a coincidence. We could talk about a similar game of “divisible by 4”. In that game, I would have 1 to 3 odds estimation that the number is divisible by 4, e.g. 1/4, and I would need a bet of 4 to 1 payout ratio or better before taking a bet that the number is divisible by 4.

  188. EnlightenmentLiberal says

    Typos, typos everywhere. Fix:

    and I would need a bet of **3** to 1 payout ratio or better before taking a bet that the number is divisible by 4.

  189. frankgturner says

    Yes P(Innocent) + P(Guilt) is not required to equal 1 because P(Innocent) does not equal P(not guilt). I can understand how someone might THINK that it does.
    .
    Just to get you thinking EL (not so much in numbers, but word definitions and interpretations) non innocence and non guilt CAN and DO overlap. A person can be considered not guilty but not innocent either. You kind of have the right idea when you talk about scenarios and what is given. I am guessing that you have heard of this before.

  190. John Iacoletti says

    There is no gambling proposition here — it’s a binary question:

    – do you believe the number of gumballs is even? yes or no

  191. John Iacoletti says

    What do you mean with your above quote? What is this difference? Can you describe a specific example please?

    People aren’t “ideal” jurors, they are human beings who aren’t always logical and without bias. That’s why they spend so much time impaneling the jury. I might be 99% confident that a person is guilty, but for reasons that have nothing to do with the evidence that was presented at trial. I should vote not guilty. In an extreme example, I could actually have witnessed the person committing the crime, but if the prosecutor doesn’t call me as a witness or present any other convincing evidence, I should vote not guilty.

  192. EnlightenmentLiberal says

    @Frank
    I don’t know what else to say. I have explained your errors – in short, you’re confusing “not guilty” as a description of the event space and “not guilty the verdict” which is a description of confidence levels. You’re seemingly hopelessly confused, and wrong.

    @John

    There is no gambling proposition here — it’s a binary question:
    – do you believe the number of gumballs is even? yes or no

    We don’t live in a world of absolutes. We live in a world of epistemic probabilities. It is a gambling proposition. Every claim you make is held to some degree of certainty, and every decision you make is a gamble based on your estimations of the truth of your premises. Every belief you hold has a degree of confidence attached to it from 0% to 100%. I might answer “Yes i believe that the number of gumballs is even”, but that’s an incomplete answer. It’s sufficient for most purposes. Most people will hopefully infer that I do not hold the belief beyond all doubt, but rather I hold it to some specific degree of confidence, such as around 99%. I think you have a problem with believing that you have absolute confidence, when you really should accept that you only have some non-absolute degree of confidence, and that you have different levels of confidence for your various beliefs.

    About the difference. Could you agree to the following? “A juror does not vote according to his beliefs that the person is innocent or guilty. Rather, a juror votes according to his hypothetical beliefs in a hypothetical world where the specific evidence is restricted to the evidence presented by the prosecution and defense.” Is that fair?

  193. EnlightenmentLiberal says

    @Frank
    Let me try it like this.

    Can we agree for the moment to ignore the possibility that the proposition of guilt is ill-defined or nebulous? I agree that’s an important consideration in the real world, but that’s just complicating matters I think. Or are you basing your entire objection off that? I think not, but I’ll ask explicitly.

    This is not a contrived assumption. Plenty of real murder trials have an extremely unlikely third option. Imagine this: “Oh your honor. I was holding the knife when I stabbed him, but it was an accident. I tripped and stabbed him. 47 times. I trip a lot.”

    This is a reasonable working assumption. It includes a lot of working assumptions, such as invisible pixies are not messing up my results.

    Under that working assumption, do you agree that the event space in the real world for a criminal trial can be accurately modeled as { “guilty”, “innocent” } ? Am I missing anything? I don’t know what it means to disagree. It’s incoherent.

    Can we talk about the sub-space of events described with the English “not guilty” ? What would that look like? In set-builder notation, I say it looks like:
    The set of “not guilty” = { X in event space : X is not “guilty” } = { “innocent” }
    Can you agree to this? If someone is actually not guilty of a crime, that means they’re innocent of the crime, right? I strongly suspect you’ll disagree, and I don’t know what to say. If you disagree, you’re speaking nonsense, and almost certainly confusing “not guilty the event space” with “not guilty the verdict”. Of course someone who is actually not guilty is innocent. That’s what words mean. There’s no third option. It’s like saying that there’s a third option to even or odd for the number of gumballs (assuming whole gumballs). If you don’t agree, then you’re not speaking English. You’re speaking some other language, and I don’t know what the hell you’re saying.

  194. EnlightenmentLiberal says

    @Frank
    PS: Again, please see the difference between “event space” and “my estimation of the truth of the matter, and my beliefs about the truth of the matter”. The event space is { “innocent”, “guilty” }. Your belief about the event can be anything in the range: “highly confident innocent”, “leaning innocent”, “undecided / agnostic”, “leaning guilty”, “highly confident guilty”, and everything in between. A “not guilty verdict” includes the estimations { “highly confident innocent”, “leaning innocent”, “undecided / agnostic”, “leaning guilty” } and excludes “highly confident guilty”. Again, you are confusing the event space – how the world actually is – with your knowledge of the world. The map is not the place.

  195. John Iacoletti says

    A “not guilty verdict” includes the estimations { “highly confident innocent”, “leaning innocent”, “undecided / agnostic”, “leaning guilty” } and excludes “highly confident guilty”.

    Exactly. Which is why the courtroom analogy is used for theist/atheist. “Atheist” includes everything except “highly confident that some god exists”. Some people want to claim that “not guilty” is the same as “highly confident innocent”, but it’s just not.

  196. frankgturner says

    EL
    Are you so dogmatically committed to the definition of Innocence being the absolute equivalent of non guilt that you are unwilling to entertain alternative interpretations?
    .
    If that is the case then I can see why you think Matt was leading people down a path of bad thinking. I would also see why you can’t seem to comprehend where others are coming from and why they neither understand you not you them. In you are going to be inflexible in how you think words should be defined and insist that people use those words in the way that YOU think is correct, then you are being no different from the theists Matt talks to, just about a different principle.
    .
    English is NOT a static language where words have a set in stone meaning that never changes. Words have meaning by consensus and if you refuse to flex with the consensus then you are doomed to being misunderstood. If you want to be a better communicator like you say that you do you need to let go of such inflexible word definitions. I don’t think you want to be a better communicator, you just want others to conform to your way of communicating. There has to be some give and take.
    .
    I HAVE ready about situations, MANY of them, where a person is not considered guilty of a crime, but not innocent either. If you really want to communicate better let go of the idea of this being impossible. You have heard of this too (at least in the USA), maybe you just disagree with the person not being considered innocent. Most lawyers will say that the person is not declared innocent and I really am wondering how you are missing this. Most people have heard of this in courtrooms (maybe you are not from the USA originally). Specific pleas can be made regarding this.
    .
    I can sympathize with you though, I used to insist that words had definitions like that and make the same accusations about people not speaking English. Maybe that is why I’ve been more patient with you than most.

  197. frankgturner says

    John does a good job of pointing out why that analogy is used. I went off on a different tangent based on my understanding of law (i am not an attorney mind you).

  198. frankgturner says

    @ EL
    I appreciate that you did ask explicitly so I will see if I can answer. To some degree guilt is I’ll defined depending on your perspective. It does take a lot of assumptions into account. I will point out that it is complicated. It does go beyond a simple mathematical claim of the event space being {guilty, not guilty}.
    .
    There are instances where a person can be considered “not guilty” but is also not innocent either. If you think that this is a trifling detail then I would like to ask why.

  199. corwyn says

    @EL:

    The event space is { “innocent”, “guilty” }.

    No, it’s really not. The event space includes things like “should not be criminalized”, “no criminal intent”. Check out some of the euthanasia verdicts for examples of this. These fail into the general category of “not guilty”, but are explicitly NOT “innocent”.

  200. EnlightenmentLiberal says

    @Frank
    Ok. I’m going to assume for the moment that you think stuff like “not guilty by reason of criminal insanity” are distinct from “innocent of the crime” and “guilty of the crime”. We’re making progress.

    Back to square 1.

    Do you agree that a piece of evidence should incline you towards a hypothesis if and only if the hypothesis predicts the evidence more strongly than all alternative hypothesis? Related, can you agree that a piece of evidence is not evidence for a hypothesis if all competing hypotheses also predict the same evidence? Then, can you agree that it’s flatly irrational to consider the question “Is the person guilty of the crime” without concurrently examining the question “Is the person innocent of the crime”, and “Is the person not-guilty of the crime by way of criminal insanity”, and so forth? If you’re with me thus far, then great. I think that’s most of my contentions.

    Suppose someone says that they have evidence that their regular garden-variety hammer is a natural object, and also says that it’s simply impossible to ever use evidence to ever demonstrate that something is supernatural. Can we agree that they are holding unfalsifiable beliefs? Can we agree that unfalsifiable beliefs of this kind are wrong-headed and pseudo-scientific? If “natural” and “supernatural” are an exhaustive listing of the physical possibility space, can we agree that any criteria for demonstrating something is natural automatically implies a set of criteria for demonstrating that something is supernatural? If you’re with me thus far, great.

    Then read this. Matt wrote this up-thread.

    I’m sorry, but you’re just wrong in saying this…in part, because we’re NOT examining a question, we’re examining CLAIMS. The claim “some god exists” can be examined even if the claim “no god exists” cannot.

    That’s simply wrong. It’s wrong for the reasons I just stated. Part of the process of examining the claim “some god exists” necessarily entails examining all plausible counter-claims.

    I have seen you Frank say similar things, and you would be wrong if you said something equivalent. This is the example of bad reasoning which Matt Dillahunty is promoting. He’s promoting unfalsifiable beliefs. Unfalsifiable beliefs are a kind of magical thinking. The falsifiability criteria is another way to look at the consequences of Bayesian reasoning that you have to examine all competing claims in order to examine a single claim. In other words, it’s impossible to examine claims in a vacuum. You always examine claims in contrast to other claims. That’s what the falsifiability criterion should teach you, and that’s what Bayesian reasoning should teach you. It seems that Matt still doesn’t understand this, and I don’t know if you do either.

  201. EnlightenmentLiberal says

    PS: And once you get that far, I think a kind-of-positivism necessarily follows. The idea that the only meaning (quote unquote) that one can assign to empirical claims is in contrast to other empirical claims – that sounds a lot like positivism to me.

  202. corwyn says

    @EL:

    You think I’m being an asshole

    No, I actually don’t. You lamented that you weren’t able to convince people, and others complained about your tone. I just tried to make that visceral to you, so that you might change your tone, if you were actually interested in reaching people.
    There was no ill intent on my part. I was trying to help. Still am for that matter.

    “O wad some Power the giftie gie us To see oursels as ithers see us!”
    — Robert Burns

    Accept the gift or not, as pleases you.

  203. corwyn says

    @John

    People aren’t “ideal” jurors, they are human beings who aren’t always logical and without bias. In an extreme example, I could actually have witnessed the person committing the crime, but if the prosecutor doesn’t call me as a witness or present any other convincing evidence, I should vote not guilty.

    You have moved here from ‘not ideal’ juror, to juror committing a crime (probably perjury). If you witnessed the crime, you can not be a legal juror.

  204. corwyn says

    @EL:

    …says that it’s simply impossible to ever use evidence to ever demonstrate that something is supernatural.

    Here is another way for you to look at Matt’s argument. Is it possible to ever use evidence to demonstrate that something has no evidence related to it? The act of finding evidence changes the perception of the thing.

    This is, I think, what Matt means by never seeing evidence for (and that ‘for’ may be particularly ‘in favor of’) the supernatural. Are there any physicists that consider that dark matter or dark energy might be supernatural? No. Even with something as nebulous as that, that we only infer because of the effect it has on natural thing, isn’t evidence that it is supernatural. No amount of searching for it and not finding it will ever convince any of those scientists to even consider that it might be supernatural.

  205. EnlightenmentLiberal says

    @corwyn
    You admitted to trolling me. I am not going to have a conversation about tone and proper discourse in this place with you. You have absolutely no cred left with me on the topic of proper tone. An apology would be a good start to redeeming some respect.

    The act of finding evidence changes the perception of the thing.

    Yes… This sounds more philosophical, and/or more like neuroscience and psychology. I agree that people’s biases change over time. I agree that beliefs and knowledge color people’s understanding and perception of many things.

    Is it possible to ever use evidence to demonstrate that something has no evidence related to it?

    I think yes. It depends on what you mean. Carl Sagan once infamously said that absence of evidence is not evidence of absence. Taken at face value, that’s wrong. Sometimes absence of evidence is not evidence of absence, and sometimes it is evidence of absence.

    For a concrete, every day scholars do surveys of the existing literature to create studies or meta-studies. I think it’s fair to characterize particular instances of this as gathering evidence to demonstrate that there is no evidence.

    This is, I think, what Matt means by never seeing evidence for (and that ‘for’ may be particularly ‘in favor of’) the supernatural.

    That would be great if that’s all he said. I would be wrong if that’s all he said. In the real world, he has also said that it’s procedurally impossible for the scientific method to show that some thing is supernatural. Which means his beliefs that some thing is natural is unfalsifiable, which means it’s magical thinking and pseudo-science. It’s just another way of describing the Bayesian approach of always comparing one proposition to its alternatives. It’s just another way of describing critical thinking – what does it mean to think critically? It means to think “What if I’m wrong?”. That includes “How would I know if I’m wrong?” and “How can I be wrong?”, and that entails the need to investigate all competing propositions. It’s all the same shit, and Matt is getting it flatly wrong in some cases.

  206. frankgturner says

    @ EL
    I think I am starting to see where you have a misunderstanding. I still think that you need to work on your tone though regardless of the source of the criticism, but I will leave that for another topic. (i am also trying to look up an old lecture or one like it in which a man used Venn diagrams to describe how the court views innocence and non innocence vs guilt and non guilt as it relates to t g e courtroom analogy).
    .
    One of the principles of science is the concept of tentative acceptance. This is accepting a hypothesis (which with enough support becomes a “theory” as I hope that you know) as the best explanation until evidence shows otherwise as the evidence supports that explanation. Matt HAS mentioned this several times, particularly with regard to evolution. I have heard him apply that to other things so I am sure that he would apply that to natural vs supernatural as well.
    .
    I see where you are going with the idea of claiming that if you cannot find evidence for the supernatural, I.e.: not natural, beca use the moment that you provide evidence it is , in effect, natural, is an unfalsifiable claim.
    .
    This is NOT what I have claimed it will claim by proposing the idea that it is impossible for science to procedurally show evidence for the supernatural. That gets more into the idea that the one dogmatic aspects of science is that it is based on observation and a e need to be able to observe things procedurally.
    .
    I am not sure what Matt’s answer would be but you could ask him rather than insisting that he has implied this and you may want to make your point about the claim implying something that is unfalsifiable. I have had conversations on this topic before though.
    .
    It is a tough issue because, in a way, one is trying to falsify the priciple of falsifiability. I tend to think of the word “natural” as being a synonym for “existent.” to say that something is “supernatural” is to say that it is “outside of” or effectively “non” existent. So if you have evidence for something being capable of being outside existence existing, well then it exists so it is not outside of existence, right? It is a bit of an oxymoron, a contradiction in terms. (I may use the words “natural” and “existent” interchangably as well as “supernatural” and “non existent”).
    .
    I have heard it said that there is hypothetical evidence for which there is no evidence, namely the imagination. We can imagine non existent things, which means that hypothetically the existent, the natural is fasifiable, even if no evidence comes along for this. Since a tremendous amount of evidence has arisen that demonstrates our existence, I would say that evidence for the supernatural is pretty slim (which is to say, there is none). And if absence of evidence is not evidence of absence, absence of evidence of an event can sure as hell make the potential likelihood of an event so small that its probability of occurrence cannot be calculated. One might even say that it is impossible to calculate the Probability of observing an unnatural or non existent event as we have an incredibly large, perhaps infinite number of observed “existent” or “natural” events and no supernatural events. Trying to calculate the Probability of an unnatural event occurring is, effectively, dividing zero by a number approaching infinity. The number on top could hypothetically be more than 0, we just have never seen it be anything but zero.
    .
    Being a good skeptic though, I am opened to other evidence (I would think Matt is as well). So it is not really that one cannot provide ANY evidence, just that we think the likelihood of them doing so is too small to be considered.
    .
    In life we have to accept some axioms, we have to make some assumptions that we cannot prove because as far as we know they are falsifiable only hypothetically. So maybe it is not that there can be no evidence for the supernatural, more that since having evidence of something makes it natural, it is held as self evident that there could not be evidence for the supernatural.
    .
    Mind you, this is based on my understanding of “natural.” Others propose definitions that are different from this that allow for a greater degree of falsifiability (which is easy, when my number is nothing). I considered talking about this actually.

  207. Kudlak says

    @Narf #141
    If they believe that the soul is eternal, unchanging and supernatural in nature what possible use is it for them to raise the point about how natural energy cannot be created or destroyed?

  208. Narf says

    You seem to be under the impression that they’ve thought through these things and have constructed coherent, consistent ideas, Kudlak.  😀  If they held themselves to that standard, they wouldn’t be theists anymore.  This is just bullshit they come up with to help them sleep at night, when they start thinking about death.  It just has to serve well enough as a metaphorical thumb to suck.

  209. EnlightenmentLiberal says

    One of the principles of science is the concept of tentative acceptance. This is accepting a hypothesis (which with enough support becomes a “theory” as I hope that you know) as the best explanation until evidence shows otherwise as the evidence supports that explanation.

    I know what you’re getting at, but I disagree with the phrasing. It should read: One of the principles of science is the concept oftentative acceptance. This is accepting a hypothesis (which with enough support becomes a “theory”) as the best explanation for as long as the evidence strongly supports that hypothesis (compared to other hypotheses).

    This is NOT what I have claimed it will claim by proposing the idea that it is impossible for science to procedurally show evidence for the supernatural. That gets more into the idea that the one dogmatic aspects of science is that it is based on observation and a e need to be able to observe things procedurally.

    This has a confusing grammar structure. I’m not sure what you’re trying to say. Let me guess. Elsewhere, you say that “supernatural” has the connotation of “not observable”. Sigh, again, not everyone uses the word in the same way, and as an effective communicator you should know that. It’s like the word “spiritual”. It’s toxic. It’s used in so many different ways that you shouldn’t use the word, except and unless a strict definition is provided in the conversation.

    Also, as I’ve said many times, if you want to define “supernatural” as “not observable”, then the concept of intrinsic methodological naturalism is completely identical to and redundant with the concept of falsifiability and other core concepts of science. Whereas, in the past Matt has clearly meant for intrinsic methodological naturalism to apply to certain observable claims, such as few months ago on show where he used Thor the thunderbolter as one such example. Further, I can quote many respectable scientists and scientific organizations which hold to this wrong position, such as the US National Academy Of Sciences, Eugenie Scott and the National Center For Science Education, and more. I can also quote seminal court cases that hold to this wrong opinion, such as the Kitzmiller v Dover trial. Thus again, most people use the phrase “intrinsic methodological naturalism” in a way other than what you mean, and I think for you to use it in this way is equivocation, confusing, and intellectually dishonest. I guess that you are trying to redefine words in order to rescue an obviously doomed principle. Don’t fall prey to a sort of ad hoc approach of saving the doomed principle (IMN). Just let the principle die the death it deserves.

    And again, even if we adpot your IMHO obscure usage of the word “supernatural” as “not existing”, IMN becomes a near-vacuous concept. If we play word taboo, IMN becomes: “It is proceedurally impossible to use science to learn about things that don’t exist.” That’s 1- obviously true, and 2- it’s rather foolish to assert it as a profound or important principle. Again, I think it comes back to your desire to save an obviously wrong principle (IMN) from falsification via intellectually dishonest ad hoc maneuvers. Just let the principle die.

    Trying to calculate the Probability of an unnatural event occurring is, effectively, dividing zero by a number approaching infinity. The number on top could hypothetically be more than 0, we just have never seen it be anything but zero.

    This is another problem I have with Matt’s epistemology. It’s IMHO best explained in the Bayesian reasonign framework, but I can also do a simple counter-demonstration. What you just wrote is an argument against any and all new physics, and cosmologists and string theorists should just give up and go home. Your argument applies just as well to study of new physics as it does to study of the supernatural. Of course we can study new physics, and obtain evidence that can convince us that new physics evidence. Your argument (and Matt’s similar argument) is complete bullshit.

    In Bayesian terms, it’s highly foolish to have a 0 prior. In fact, it’s highly foolish to have priors that are themselves not informed by evidence – give or take your starting 50 50 priors (and/or a careful analysis of Kolgomorov complexity of the hypothesis in order to create your prior, or some other similar process). A 0% prior is identical to being dogmatic. A 0% prior is identicail to proclaiming that you will dismiss any and all evidence to the contrary.

    To compare and contrast: “Extraordinary claims require extraordinary evidence”. When reasoning properly, we should conclude that claims are extraordinary when we have mounds of previous evidence against the claim. For example, it’s an extraordinary claim to claim that I can fly by flapping my arms (absent any machinery, technology, etc.), and it’s an extraordinary claim precisely because of the wealth of evidence I have against that proposition. That’s why it requires a comparable mound of evidence in favor of the proposition before I would accept it (not comparable by volume per se, but comparable in compelling-ness).

    <blockquote
    In life we have to accept some axioms, we have to make some assumptions that we cannot prove because as far as we know they are falsifiable only hypothetically. So maybe it is not that there can be no evidence for the supernatural, more that since having evidence of something makes it natural, it is held as self evident that there could not be evidence for the supernatural.

    Further down the rabbit hole you go. You just said that you are a dogmatic atheist and that one of your axioms is to assert that there is no god. At the very least, that’s what most religious people will understand when they read what you wrote. Do you understand this? This is precisely what I complained about w.r.t. Matt’s bad reasoning, and now you just embraced it. It’s like a perfect reductio ad absurdum in favor of my position.

    That, or you claim that an existing god is natural a claim I’m conditionally ok with, but you need to preface that a lot before you make the claim precisely to avoid having theists think that you are dogmatically opposed to the existence of their god.

    I suppose I should ask, are you dogmatically opposed to the existence of the Christian god? I don’t care right now if you want to call it natural or supernatural. Right now, is it proceedurally coherent that science could show that the Christian god exists? Right now, can you imagine hypothetical evidence which would make you tentatively but strongly convinced that there is a powerful creature (in space time) that goes by the title “Christian god” which can change any sort of observable aspect of reality into any reasonably consistent and coherent state, and seemingly by force of will? I can. If you can’t, then you’re not trying hard enough. I know Matt is famous for saying he cannot imagine it, and he’s not trying hard enough either. Just imagine the plot of any popular Hollywood movie. Don’t think small like the stars rearranging to spell out “I am here”. For example, try Bruce Almighty. It’s honestly not that hard to think of evidence that should convince you.

    PS: Of course, the evidence that would convince me is so utterly unlike anything proposed by Christians and other religious people. My required evidence is so completely over the top in its strength compared to the usually proffered “evidence”. I’m sure many Christians might complain that it’s unreasonable to expect evidence on the scale of Bruce Almighty. My result is “tough shit”. This is a result of the completely dismal state of affairs of the Christians and their research program, and the mounds of evidence against their god. The only evidence the Christians can attempt to proffer is nowhere near enough to the threshold of convincing, and that’s because their god does not exist. If their god did exist, then we would have evidence like Bruce Almighty. (IIRC, Richard Carrier makes similar arguments in his book Why I Am Not A Christian.)

    PPS:
    I think this also goes to a sacred cow of hosts of the show. It’s the idea that we can wean people off religion by showing that the evidence should not convince them. I’m not saying it’s doesn’t work, but I would prefer more rigorous thinking. I also expect that my way would convert more people.

    The idea that you can wean someone off their religious beliefs without offering concrete alternatives is IMHO bumcus. If the religious person believes that there are no plausible alternatives, whether rightly or wrongly, then no amount of “but that evidence shouldn’t convince you” should change their mind. In their position, they see only one plausible scenario, and no amount of attacking the evidence will change that result. It’s a simple consequence of modeling it with Bayesian reasoning, and any good Christian thinker who is ignorant of alternatives is rationally obliged to stick to their guns no matter how flimsy the evidence. In order to get them off their beliefs, you need to propose that alternative, even if that alternative is “We don’t know yet”. As the famous Sherlock Holmes once said, if you elimate the impossible, then whatever remains, however improbable, must be the truth. (Which is itself a simple consequence of Bayesian reasoning.)

    IMHO, we should focus on showing alternatives where we have them, and emphasizing the difference in the evidence between their bad proposition and our right proposition.

    For religious propositions where we don’t have a secular alternative (yet), we do have to adopt the approach of the show and try to argue them down to “I don’t know, and you don’t know either” by attacking their evidence, but again IMHO this should be the option of last resort, not the first.

    I know that what finally convinced me of the false-ness of the world’s religions was not coming to the realization that my evidence was crap. What finally convinced me was the realization of a positive alternative, specifically: all of those other religions must be made up, and there must be simple non-religious explanations for how those religious came about, and whatever explanation works for all of those other religions must work for mine.

  210. frankgturner says

    @ EL
    The problem is the usage of one fuzzy intangible concept that individuals define in their own way and no one really has a solid factual definition for, “supernatural,” to explain ANOTHER fuzzy intangible concept that no one has a solid definition for, “God” (ANY God, not just the Xtian one). It’s is like trying to bake cake with unknown portions of mud and egg shells.
    .
    Words have meaning by consensus. The general consensus that I have ascertained from this group largely equates the word supernatural with existence. I don’t know what YOU personally mean by it. Can you give your definition?
    .
    I will respond and reply to more in a bit when I have time. I will point out that you are seeming less like an atheist or agnostic and more like a believer in a state of doubt, but I don’t know.

  211. EnlightenmentLiberal says

    @frank
    There is enough of a consensus of the meaning of the word “god” to claim that there are no gods. At the very least, a god is a powerful creature which exists and which accomplishes amazing feats seemingly through force of will. We know no such thing exists.

    Most people do not understand the word “supernatural” to mean “does not exist”. Most people understand “supernatural” to be non-identical to “does not exist”. You are using words contrary to their normal usage.

    Even if we grant your usage, the claim of IMN becomes vacuous, and thus wrong-headed. IMN becomes “science does not work on things which do not exist”. Well, no shit. It’s not some profound realization. However, if you then argue that IMN means science cannot help us learn about gods, then you’re committing an equivocation, or admitting that it is a dogmatic faith belief that there are no gods – a horrible position to hold.

  212. frankgturner says

    EL
    You do such a good job of putting words into people’s mouths and claiming that they said things that they did not say that I wonder if you really are reading their responses. Maybe to be a better communicator you ought to listen to what people ARe saying and not so much what they are not saying. You are trying my patience and I am not far from considering you not worth talking too. Which is too bad as you make some good points and are a pretty solid guy as far as intellectualism.
    .
    I will humor you as you give a solid definition for a God which is more than one gets from a lot of people. What is your definition of “supernatural” and what do you feel there is a consensus on regarding that definition?
    .
    P.s.: You might want to be a bit less accusations of you don’t want to come off as an asshole. (Of course I think that if I stop talking to you that you will just change nicks again and come on here under a different name, similar to what you did in the past. I was the one who showed you the most patience before though).

  213. EnlightenmentLiberal says

    @Frank

    You do such a good job of putting words into people’s mouths and claiming that they said things that they did not say that I wonder if you really are reading their responses.

    I’m being dead serious here that it might help if you point out specific examples where you feel that I have done so. I do not know what you are talking about.

    What is your definition of “supernatural” and what do you feel there is a consensus on regarding that definition?

    To answer your question, I am going to make a prediction. Let’s imagine we gave a survey to a large number of Christian believers. Simple random sample. Survey has the following questions:

    – Question: Does god exist? yes / no

    – Answer this question only if you answered “yes” to the earlier question. Question: Is god supernatural? yes / no

    In this hypothetical scenario, I make the following prediction: 90% or higher of the survey answers for question 1 will be “yes”, and 90% of those who answered question 2 will be “yes”.

    As you should well know, I happen to think that the word “supernatural” is bullshit, underspecified, and nebulous., in exactly the same way that “spiritual” is bullshit, underspecified, and nebulous.

    However, I also recognize that many people have this irrational attachment to the idea that this label is a meaningful concept and that it applies to their god hypothesis. It is bad communication and intellectually dishonest to pretend that this word has some other usage. It is bad communication and intellectually dishonest to speak as though “supernatural” is identical with “non-existent”.

    P.s.: You might want to be a bit less accusations of you don’t want to come off as an asshole. (Of course I think that if I stop talking to you that you will just change nicks again and come on here under a different name, similar to what you did in the past. I was the one who showed you the most patience before though).

    Wait what? I changed names once, long ago, for personal reasons, and not because people stopped talking to me. I’m well past caring that people think I’m an asshole for standing by my assertions and defending them forcefully. I’m definitely willing to back off of my accusations when I stop being convinced that they are substantiated, but not a moment before. I’m going to make accusations all I want if I think that they are honest and substantiated, and you better well deal with it.

    If you want to make progress on that front, we can talk about your recent accusation where you say I have put words into someone’s mouth. Was it your mouth? What words? I’m not a mind reader. I don’t know what you’re talking about. Obviously I don’t understand something of your position, and I’m not going to understand it when you just make vague references to some purported strawman.

  214. EnlightenmentLiberal says

    It is bad communication and intellectually dishonest to speak as though “supernatural” is identical with “non-existent”.

    Specifically, I’m willing to entertain arguments that synthetically it is true that “supernatural” is identical with “non-existent”, e.g. I am willing to entertain the notion that it is an empirical fact that these two distinct concepts happen to name the same set of things in reality. In fact, I think that by most of the reasonable definitions of “supernatural”, it is a synthetic fact e.g. empirical fact that “supernatural” and “non-existent” name the same set of things in reality.

    However, it’s intellectually dishonest and bad communication to pretend that these words have the same analytic meaning, e.g. the same definition.

    The truth of the assertion that supernatural things do not exist is a conclusion of the available evidence and science. It is not something that we can gleam with the cliche sitting on a chair without looking at the world.

  215. frankgturner says

    @EL
    And very well on the part of synthetic fact. I think the difficulty comes from one term being so nebulous that what it winds up describing is ultimately either non existent or not relevant. I can;t really sy for sure but hopefully you are starting to get the original point that I am making.
    .
    All right let me see if I can get you to get this. A LOT of this is about TONE.
    .
    You accuse Matt of being wrong headed in making the assertion that there can be no evidence for the supernatural, right (if I am getting you)? Yet the term is so nebulous and undefined that you can’t really define it. Can one have evidence for something that is nebulous and undefined? Perhaps. Maybe one can have evidence for the more solid and concrete parts of the nebulous word or idea, if supernatural even HAS any concrete parts to it. Does that really count as evidence for the word or idea, for the supernatural? (You yourself say the word is bullshit). A good skeptic might say, “I don’t know, but I suspect that…”.
    .
    You didn’t, you forcefully defended the proposition that Matt believed in an un-falsifiable position. In context you may not have meant that this was hypothetical (like perhaps you don’t understand what he is getting at given that he is using nebulous terms). What it came across as though is that you were certain that he did. It is not like he said that he NEVER would believe that ANY evidence could prove the supernatural, which is a bullshit term ANYWAY. You asserted forcefully though that it was bad thinking.
    .
    That’s interesting because you talk about eliminating all other hypotheses to come to the truth, yet you did not seem to eliminate the possibility that maybe, given the fact that he was using a nebulous term like “supernatural,” he was expressing ANOTHER hypothesis that did not equate to belief in an un-falsifiable position. I do not interpret the claim that, “you cannot provide evidence for the supernatural,” to mean that the natural is un-falsifiable. I basically see it as, you can’t provide evidence for bullshit. If you have hard evidence for something, it isn’t bullshit! Do I believe that dogmatically? No, because there is a wide range of ideas for what constitutes bullshit and it is ULTIMATELY a matter of opinion. So trying to make the claim that “you cannot provide evidence for bullshit,” implies un-falsifiability of the complement is sort of goofy, because what is the complement of bullshit?
    .
    I DID think of one hypothesis, specifically with regard to the terms “supernatural” and “non-existent” being synonyms (though not precisely the same, similar). Hence why I discussed the principle of evidence (which exists) to prove the “non existent,” for which there cannot be evidence. I was not stating this dogmatically, I was merely using it to make a point. It was a hypothesis about what he meant. I don;t know exactly what he meant because I can’t measure his brain waves or get a demonstrable set of data about what he believes. It seemed to make sense to me because providing evidence for something that cannot have evidence to support it seems like a pretty good starting axiom, particularly when talking to theists who are trying to claim the opposite.
    .
    Yet I was accused of equating the two terms, FORCEFULLY, by you, as though I believed that they were perfectly identical. I don’t, though I think they are synonyms (SIMILAR in meaning). You did not ask me if maybe I was speaking in more hypothetical terms? I thought I made that abundantly clear, but even if I didn’t, I don’t think I pushed that issue forcefully enough to say, “I am absolutely and in no uncertain terms convinced that supernatural and non existent are exactly the same terms.” I said that i may use them interchangeably (for the sake of argument at the time).
    .
    And then, despite making it what I thought to be abundantly clear that I was trying to make a hypothetical equivalency in order to provide a more coherent argument NOT based on a bullshit term. you accuse me of being intellectually dishonest for pretending, yes PRETENDING (for the sake of argument in this case, which I thought would help make ideas clearer) that they are the same? People do that all the time! If you had said to me, “are you trying to claim that these have perfectly identical meanings?” No, I immediately get accused of being intellectually dishonest. Had I been an average person my response would have been a great big “F*** you!” You didn’t ask, you did not even THINK to ask. You went RIGHT into accusation. (I did not take it personally but others would, which is why I am pointing this out).
    .
    So you basically see how defending your assertions too forcefully can sound, well TOO assertive and like you are convinced of your accusation in absolute terms (or semi-absolute). Yet when YOU developed a feeling about someone ELSE defending themselves forcefully (like corwyn in 108) you accuse HIM of acting self righteous. An no one is allowed to accuse YOU of acting self righteous? Yes he admitted to sour intentions (trolling), but was it really? Was it really trolling given that he was merely copying your phrasing, doing a bit of mirroring?
    .
    With all due respect YOU have sounded more than a little self righteous and I have a hard time believing that all of your intentions were genuine. I have no proof of this, so I will just point out that it is a feeling. I would ask you but I am not sure that it is worth it at this point. If I did it would demonstrate a point though, when terms are nebulous and undefined or something is obviously a matter of opinion that cannot be measure empirically, ASK. The person may not give you a completely honest answer but that might be the best that you can get. (In actuality I don’t think you are trying to sound self righteous, but at times I don’t think you give it serious consideration either.
    .
    You try to apologize for bullying and say you are sorry if people felt bullied, but I never saw you asking EXACTLY what might have made people feel bullied. THAT might be a good way to show some considerations for people’s feelings. Well defending your OPINIONS too forcefully is one way to come off as self righteous, intentionally or not, and to behave like a bully, intentionally or not. You really need to start asking what made people feel certain ways and CARE what they have to say if you want to not offend them. I know you say that your way of weaning people away from religion may work better, but if it came from you being as inconsiderate about people’s feelings and as offensive as you appear to be and offended them, I doubt that they would give a shit. So coming from you it would not be effective.
    .
    I do like some of what you have to say but you make a LOT of presumptions about other people’s positions that you may have some evidence for, but you most certainly have NOT burned away all of the possible hypotheses about what they are saying to make accusations with the degree of confidence as you do. It is appearing to come off as self righteous and you might make a habit of ASKING if you have correctly interpreted what others have to say. Have you ever heard of the psychological exercise of reflection? You do this thing where before you respond to someone’s questions you re-state it in your own words, then they do the same for you. It is a really good exercise for determining if you are reading people correctly.
    .
    Ok so enough about tone.
    Just to be clear, I think supernatural is a bullshit nebulous term too. As far as,

    Also, as I’ve said many times, if you want to define “supernatural” as “not observable”, then the concept of intrinsic methodological naturalism is completely identical to and redundant with the concept of falsifiability and other core concepts of science

    I don’t think I have heard you say it many times on here or any other message board, but I presume that you may have said it. Given that this is the first time on these boards that I made a hypothetical relationship between “supernatural” and “not observable,” I don’t see how you could have said it before, at least not on here. Maybe you were having a conversation in your head. That or maybe, “As I have said many times before” is just a lead in that does not mean anything.
    .
    And I don’t think that I have talked about methodological naturalism on these boards before. I have listened to conversations about it. Were you conversing with Matt and got him confused with me? Are you putting HIS words into my mouth?
    .
    I have discussed methodological naturalism with others before and I have pointed out that one of the dogmatic tenets of it appears to be that evidence must be observable. Also, if I am not mistaken methodological naturalism USES falsifiability. So if you hypothetically eliminate some of the things that make them distinct, they do become more similar and perhaps the same.
    .
    Now THIS is an example of putting words in my mouth.

    In life we have to accept some axioms, we have to make some assumptions that we cannot prove because as far as we know they are falsifiable only hypothetically. So maybe it is not that there can be no evidence for the supernatural, more that since having evidence of something makes it natural, it is held as self evident that there could not be evidence for the supernatural.
    Further down the rabbit hole you go. You just said that you are a dogmatic atheist and that one of your axioms is to assert that there is no god. At the very least, that’s what most religious people will understand when they read what you wrote. Do you understand this?

    I was going, wtf? THAT is totally not where I was going.
    I was getting that from The Republic (paraphrasing). The idea is that there are some things that we have to accept without proof. PLENTY of religious people accept God without proof. That is effectively taking god as an axiom. Many will even argue that God is beyond proof (I think it is bullshit but there you go).
    .
    Some religious people could take it the way you are describing when I think about it. As you argued about what I need to preface things with given the way people might take it, i agree. I also think that you are the pot calling the kettle black. (I.e.: you need to do the same thing, BIG TIME).
    .
    The main principle behind the natural or supernatural something goes like this. Natural things are created by nature. Now man made things are created by our will and hard work, like cars. However, we are part of nature, cars are part of nature. So, in a way, cars are natural (maybe not naturally made, but they exist within nature and are made by a being who was made by nature and exists within nature). THAT is why I, for the sake of discussion and to try to make a nebulous term less nebulous, equated the words exist and nature and non existence and supernatural. It is not a perfect synonym, supernatural could be “above” natural or “beyond” natural. However, if something is beyond nature, could not we expand the sample space (to use a mathematical term) to just include the area beyond? So with regard to the larger sample space, the something beyond nature, like a god, is still a part of nature? So, in a way, EVERYTHING that exists is natural.
    .
    For a cute analogy, I think of George Carlin making the remark that the term “situation” is redundant. When the news says that “police are responding to an emergency situation,” why not just, “police are responding to an emergency?” EVERYTHING is a “situation.”
    .
    More words put in my mouth,

    IMN becomes “science does not work on things which do not exist”. Well, no shit. It’s not some profound realization

    I was not claiming that it was profound. It is an axiom. And I never said that specifically but I was talking about axioms and I CAN see where this was implied. In scientific endeavors you search for evidence. Evidence helps to prove that things exist or don’t exist. Science cannot procedurally provide evidence for the supernatural MIGHT just mean that science can’t provide evidence for that which does not exist. It is not rocket science but there ARE those to whom this needs to be stated.
    .
    How in the hell did THAT become,

    if you then argue that IMN means science cannot help us learn about gods, then you’re committing an equivocation, or admitting that it is a dogmatic faith belief that there are no gods – a horrible position to hold

    ?
    .
    I have no idea where the hell you got it that I was going in THAT direction. That was definitely putting words in my mouth as I have no idea why or hell you would have thought that was implied. You said several times about us having different conversations.
    .
    One thing about the courtroom analogy again. If you listen to law lectures they really do talk about how “innocent” and “guilty” are not the same thing. I have read many a lawyer say when a person tells them they are innocent the lawyer respond with, “I am not interested in why you are innocent, I am interested in how you are not guilty.” I have seen a lecture (for the life of me I can;t find it but there are other boards where people talk about this) where innocence and guilt are drawn as Venn diagrams. The innocence and guilt circles of a particular crime (each crime is shown as a separate diagram) are disjoint (showing that one cannot be both innocent and guilty of the same crime) but the areas outside of each circle show that the region of not guilty includes the circle of innocence, but also includes a region of non innocence and non innocence includes guilt and non guilt. So non guilt and non innocence have overlap. I figured that you had heard this or WOULD look it up (I will give a link at some point).
    .
    The other things that the lecture gets at is how innocence and guilt of one crime does not equate to innocence and guilt of another. They also talk about innocence and guilt being highly conditional (so there are lots of “scenarios,” as you mentioned before, I just didn’t get into it). What I got out of it was something to the effect of P(Innocent) = 100% – P(guilty) – ( P(~innocent) and P(~guilty) ). That is oversimplifying it as guilt and innocence are much more complicated principles that are highly conditional, I am just using it to demonstrate that as far as the court is concerned, there is overlap between non guilt and non innocence (even of a single crime). It is probably because the question of whether what happened was a crime as well as whether the action was committed by the person in question. I think I get what you are getting at though, there is a certain range of sample space.
    .
    There IS something though that i still think you are not getting. I am tired though and will have to get to this later.

  216. EnlightenmentLiberal says

    @Frank
    So, is it fair to say that you’re mad at me not because of my general philosophical (and epistemology) claims? Is it fair to say that you’re mad at me not because of the strength and forcefulness because in how I made those claims? Is it fair to say that you’re mad at me solely because you think I’m not accurately or fairly describing Matt’s position?

    So, I’ve already mentioned I’d like to avoid talking about what Matt has said as much as possible, and instead focus on the actual interesting points. However, it seems that I do have to go back to this petty discussion about things Matt said or didn’t say. Fine.

    Citations that of 1 March 2015 Matt makes a claim with the effect that the intersection of supernatural things and observable things are non-empty. Alternatively, Matt does not realize the implications of his statement, and that implication is that the intersection of supernatural things and observable things are non-empty.

    Episode 907 from the official show archive, mp3 format, about 44 min 30 seconds in. Quoting Matt:

    We cannot actually calculate the probability for this because we have no way to calculate how likely it is that someone who is actually experiencing – you know – what they claim to be experiencing. All we know right now is that there is insufficient evidence to justify the explanation that they’re giving. “Oh yes I did in fact go to heaven.” How can we verify? I just posted the video this week on the idea of supernatural existence and supernatural causation and how my assertion is not that the supernatural doesn’t exist. It’s really a recognition as science has done forever that we cannot – we do not have a mechanism by which we can confirm the existence of the supernatural or causation by the supernatural. And until that changes…

    Episode 907 from the official show archive, mp3 format, about 46 min 15 seconds in. Quoting Matt:

    So there’s this thing where we’ve got something and we don’t have an explanation for it, but we can make a statement about it, and that is “either there is a natural explanation for it, or there’s not”, and the way people get to asserting the supernatural explanations is so far as I can tell always an argument from ignorance fallacy, where we don’t have a natural explanation for this therefore it must be a supernatural explanation, and they’re avoiding the obvious answer which is we do not yet have a known natural explanation for this. I don’t know how they came to the conclusion that a supernatural explanation is more likely than all of the unknown potential natural explanations that are out there. That is how we failed in – “I don’t know where lightning comes from, but I can’t think of any explanation for it, so Thor is throwing them down.” Congratulations. It’s a fallacy.

    I agree with many things that Matt is saying here. I agree that “I don’t know where lightning comes from, and I cannot think of any explanation for it, thus god” is a fallacious argument from ignorance. I agree. However, where Matt goes wrong is when he says:

    We cannot actually calculate the probability for this because we have no way to calculate how likely it is that someone who is actually experiencing – you know – what they claim to be experiencing [EL: something supernatural].

    I just posted the video this week on the idea of supernatural existence and supernatural causation and how my assertion is not that the supernatural doesn’t exist. It’s really a recognition as science has done forever that we cannot – we do not have a mechanism by which we can confirm the existence of the supernatural or causation by the supernatural.

    Matt is pretty clear. One cannot be clearer than this.

    This is just one citation. However, I can find more that make the same points, again, and again, and again.

    Quoting you Frank:

    You accuse Matt of being wrong headed in making the assertion that there can be no evidence for the supernatural, right (if I am getting you)?

    Correct.

    Yet the term is so nebulous and undefined that you can’t really define it.

    I believe that I could give a definition that matches most common usage. However, I think the term is largely useless and it causes us to make mistakes in our thinking.

    It is not like he said that he NEVER would believe that ANY evidence could prove the supernatural, which is a bullshit term ANYWAY. You asserted forcefully though that it was bad thinking.

    Again, see citations of Matt above please. And again, I can find dozens more just like that.

    It does not matter that I believe the term is bullshit. What matters is that Matt believes the term is not bullshit, and Matt believes that Thor the thunderbolter is an example of a supernatural claim, and that Matt believes that it’s procedurally impossible to use science to confirm the existence of Thor the thunderbolter.

    That’s interesting because you talk about eliminating all other hypotheses to come to the truth, yet you did not seem to eliminate the possibility that maybe, given the fact that he was using a nebulous term like “supernatural,” he was expressing ANOTHER hypothesis that did not equate to belief in an un-falsifiable position.

    One: He gave Thor the thunderbolter as a specific concrete example of supernatural phenomena.

    Two: Many times he has implicitly or explicitly claimed that everyday mundane objects in our world, like a hammer, are natural. (Citations not needed.) In light of the earlier quote where Matt says that it’s procedurally impossible to use science to show that something is supernatural, it thus means that Matt’s belief “a hammer is natural” is unfalsifiable.

    I do not interpret the claim that, “you cannot provide evidence for the supernatural,” to mean that the natural is un-falsifiable. I basically see it as, you can’t provide evidence for bullshit.

    You are not the guarantor of the word “supernatural”. Matt uses the word differently than you. Matt expressly gave an example of Supernatural phenomena which is not “bullshit” in the sense of ambiguous or incoherent. Thor the thunderbolter is a perfectly coherent, sensible, unambiguous, and testable proposition.

    I suspect that Matt is confused on the matter and uses the word “supernatural” inconsistently, but that is no excuse or defense against my accusations that he is teachings wrong things on the show and in his other videos.

    If you have hard evidence for something, it isn’t bullshit!

    In the above quotes, Matt made two relevant assertions. 1- Either something is natural, or it’s supernatural (implicitly made). 2- It’s procedurally impossible to use science to demonstrate that something is supernatural.

    The logical conclusion from this is that it’s also impossible to use science to demonstrate that something is natural. As I have been trying to emphasize for a while now, the only way that something counts as evidence for a proposition (to a first approximation:) is if and only if the proposition predicts that evidence, and all competing propositions do not predict that evidence. To describe it in another way, if you have no alternative propositions, then you cannot have evidence for your proposition. To describe it in another way, the extent to which empirical claims like “a hammer is natural” are meaningful are to the extent they can be distinguished from competing hypotheses.

    Again, under Matt’s assertions, Matt does not have evidence that his hammer is a natural object. Matt’s belief that a hammer is natural is an unfalsifiable belief according to Matt’s assertions.

    To put it another way, if Matt cannot imagine specific and concrete evidence that would convince Matt that his hammer is a not natural object, then it’s an unfalsifiable belief.

    I don;t know exactly what he meant because I can’t measure his brain waves or get a demonstrable set of data about what he believes.

    You can get a sufficient approximation of what he believes by listening to him. You’re just not trying very hard.

    You did not ask me if maybe I was speaking in more hypothetical terms? I thought I made that abundantly clear, but even if I didn’t, I don’t think I pushed that issue forcefully enough to say, “I am absolutely and in no uncertain terms convinced that supernatural and non existent are exactly the same terms.” I said that i may use them interchangeably (for the sake of argument at the time).

    I’m sorry. I do not recall you saying “for the sake of argument” or any similar disclaimer.

    So you basically see how defending your assertions too forcefully can sound, well TOO assertive and like you are convinced of your accusation in absolute terms (or semi-absolute).

    Knock this shit off already. I’m being just as forceful as I would be with someone who says that evolution is false, but I’m not dogmatic about truth of evolution. I’m just very convinced. It’s just irritating and it’s not helping the conversation in any way, just like creationists who say that I am a dogmatic evolutionist. Calling me names (e.g. “dogmatic”), aka doing an ad hominem, is not productive. It’s just annoying.

    Yet when YOU developed a feeling about someone ELSE defending themselves forcefully (like corwyn in 108) you accuse HIM of acting self righteous.

    Notice the difference between what you are doing and what I did. I first explained to corwyn in clear simple terms how he is wrong, how he didn’t read a thing I wrote, and then I made a suggestion that perhaps he should simmer down, read what I write, in order that we might have a more productive conversation.

    Now, it’s apparent that you think I’m not doing the same service to you. I don’t think so. I think you need to work on your communication skills.

    Case in point:
    I’m sorry for going into such excessive detail, but it’s necessary.

    Recently, you wrote several things that AFAIK originally refer to when you said this:

    I tend to think of the word “natural” as being a synonym for “existent.” to say that something is “supernatural” is to say that it is “outside of” or effectively “non” existent.

    I took that at face value. You said you tend to think that, and I understood that as meaning that you tend to think that. I did not see “for the sake of argument”. I did not see “let’s consider this definition”. No. You stated a (weak) belief that this is what the words mean. There was no hedging of bets. There was no disclaimer.

    In reply to what I said about what you said, you said this:

    I DID think of one hypothesis, specifically with regard to the terms “supernatural” and “non-existent” being synonyms (though not precisely the same, similar).

    You did not describe it as a hypothesis which you are currently examining. You stated it as an hypothesis which you accept, a belief.

    Hence why I discussed the principle of evidence (which exists) to prove the “non existent,” for which there cannot be evidence. I was not stating this dogmatically, I was merely using it to make a point. It was a hypothesis about what he meant.

    Again, you were not doing a “for the sake of argument”. You stated a belief.

    Further, you have another confusion. I did not accuse you of being dogmatic about the belief of what the word “supernatural” means. I said that given you use this meaning of the words “supernatural” and “natural”, then your belief “a hammer is a natural object” is held dogmatically. Big difference.

    Yet I was accused of equating the two terms, FORCEFULLY, by you, as though I believed that they were perfectly identical. I don’t, though I think they are synonyms (SIMILAR in meaning). You did not ask me if maybe I was speaking in more hypothetical terms?

    Fuck no. I’m not going to sit here and ask “Did you really mean that? Or did you forget a “for the sake of argument” or something?”. If you say that you believe something, then I’m going to take you at face value. Rather than throw a temper tantrum that you misspoke, you could have just clarified what you meant, and we could have moved on.

    I thought I made that abundantly clear, but even if I didn’t, I don’t think I pushed that issue forcefully enough to say, “I am absolutely and in no uncertain terms convinced that supernatural and non existent are exactly the same terms.” I said that i may use them interchangeably (for the sake of argument at the time).

    I covered this confusion above. tl;dr I didn’t call you dogmatic in your belief of the word supernatural. I called you dogmatic about your belief “a hammer is a natural object” given your belief about the word supernatural. Again, big difference.

    And then, despite making it what I thought to be abundantly clear that I was trying to make a hypothetical equivalency in order to provide a more coherent argument NOT based on a bullshit term.

    Again, you were not abundantly clear. You were not even clear in the slightest. You stated a belief, a weak belief, but a belief nonetheless. Nowhere did you even hint that this is a hypothetical position for the sake of argument.

    Had I been an average person my response would have been a great big “F*** you!” You didn’t ask, you did not even THINK to ask. You went RIGHT into accusation. (I did not take it personally but others would, which is why I am pointing this out).

    Perhaps projection? I am in a bad mood about this entire thread, and I am generally not going to give you an inch that you do not deserve (or deserve with the principle of charity, etc.). You made a mistake. You thought you said “for the sake of argument” or something similar, but you did not. Deal with it. Own up to your mistake, and we can move on.

    With all due respect YOU have sounded more than a little self righteous and I have a hard time believing that all of your intentions were genuine.

    Do what you must. I am honest to a fault. I hate lying. If you think I’m lying, or if you think I have deceived myself, then I do not know what I might say to convince you otherwise. I am at a loss.

    You try to apologize for bullying and say you are sorry if people felt bullied, but I never saw you asking EXACTLY what might have made people feel bullied.

    True. I could have asked that question.

    THAT might be a good way to show some considerations for people’s feelings. Well defending your OPINIONS too forcefully is one way to come off as self righteous, intentionally or not, and to behave like a bully, intentionally or not.

    I am acting self righteously now, where I understand that to be largely synonymous with “arguing clearly and stringly and believing that oneself is correct”. Again, if you think this is being a bully, then I’m sorry to say that I will have to dismiss your feelings on this particular matter, and I will not take them into account. If all you have is “arguing clearly and strongly and believing that oneself is right” makes a bully, then I’ll just have to tell you to fuck off on this particular complaint. Again, to the extent that I’m misrepresented your positions, that’s something which we can work on.

    I know you say that your way of weaning people away from religion may work better, but if it came from you being as inconsiderate about people’s feelings and as offensive as you appear to be and offended them, I doubt that they would give a shit. So coming from you it would not be effective.

    My primary goal here is not to change people’s minds. My primary goal here is to have fun, by engaging with people honestly so that I might see if my positions are mistaken. I like learning and arguing in good faith.

    I think you need to be reminded that ideas and beliefs do not deserve respect. People deserve respect, not beliefs. Reasons for beliefs deserve respect, not beliefs. If you think it’s offensive and bullying for me to say you’re wrong clearly and without hedging, with explanation of how you are wrong, then again I’m simply sorry to say that my response will be fuck off for this particular concern. I do not care and will not care if you think I’m offensive because I’m telling you you’re wrong and how you’re wrong. Again, I suggest that you grow up, learn that other people will challenge your beliefs, that you can be wrong, and you should learn to deal with it.

    For the most part, I have been careful to attack beliefs and not persons. To the extent that I attacked the person and not the belief, I do apologize, and I will be more mindful of this in the rest of the discussion. However, I will pull no punches in criticizing what I believe to be beliefs of yours which are bad and foolish.

    you might make a habit of ASKING if you have correctly interpreted what others have to say.

    I think the central core of this disagreement is that you believe differently about some of Matt’s positions. For that reason you think I’m acting without good cause when I describe and attack Matt’s position. Simply put, on this point, you’re wrong, completely and utterly wrong. I’ve provided one citation. I can provide a dozen more if that’s what it will take.

    Also, as I’ve said many times, if you want to define “supernatural” as “not observable”, then the concept of intrinsic methodological naturalism is completely identical to and redundant with the concept of falsifiability and other core concepts of science

    I don’t think I have heard you say it many times on here or any other message board, but I presume that you may have said it. Given that this is the first time on these boards that I made a hypothetical relationship between “supernatural” and “not observable,” I don’t see how you could have said it before, at least not on here. Maybe you were having a conversation in your head.

    I’m sorry. I have fallen prey to the common English rhetorical device of using the word “you” to make a general comment. I should have used the word “one”.

    You are not the first person with which I’ve had this conversation. I have made this made many times in conversations with people other than you.

    I’ve have said that many times before on this blog. I suggest you check out this link for starters, and the comments there-in.
    http://freethoughtblogs.com/axp/2014/12/14/open-thread-for-aetv-896-matt-and-jen/

    Now THIS is an example of putting words in my mouth.

    In life we have to accept some axioms, we have to make some assumptions that we cannot prove because as far as we know they are falsifiable only hypothetically. So maybe it is not that there can be no evidence for the supernatural, more that since having evidence of something makes it natural, it is held as self evident that there could not be evidence for the supernatural.

    Further down the rabbit hole you go. You just said that you are a dogmatic atheist and that one of your axioms is to assert that there is no god. At the very least, that’s what most religious people will understand when they read what you wrote. Do you understand this?

    I was going, wtf? THAT is totally not where I was going.

    You and I both understood that I was not claiming you literally said “I am a dogmatic atheist” especially because I included the quote just above. Rather, you and I both understood that I was giving a characterization of your position and of its logical conclusions. That’s not putting words in your mouth. Specifically, this is not a strawman. This is not me falsely attributing words to you. Instead, this is me giving my characterization of your position and of the logically entailed facts about your position. Your assertion of “putting words in your mouth” is not founded.

    Rather, you are upset that I claim that your position can be accurately characterized in that way. We can have a discussion about that, but it is no strawman, and nothing dishonest has occurred.

    However, you then explain that “if a god exists, then it is natural”, which means that we’re using words in inconsistent ways. Again, it’s very difficult to have a conversation with someone who will use words contrary to their common usage without providing an explicit warning.

    More words put in my mouth,

    IMN becomes “science does not work on things which do not exist”. Well, no shit. It’s not some profound realization.

    Please explain where I said you said it’s a profound realization. I’m not seeing it. Rather, it appears that you’re reading too far into what I say, reading between the lines, and inventing things I said rather than responding to what I have actually said.

    I was not claiming that it was profound. It is an axiom. And I never said that specifically but I was talking about axioms and I CAN see where this was implied.

    Just before you said that it was just a “for the sake of argument” hypothetical. Now you’re saying it’s an axiom. Pick one asshat. (And yes, that is an attack on your person, because I’m getting sick of things.)

    I have no idea where the hell you got it that I was going in THAT direction. That was definitely putting words in my mouth as I have no idea why or hell you would have thought that was implied. You said several times about us having different conversations.

    Again, “putting words in your mouth” does not mean what you think it means. I’m drawing logical conclusions from your beliefs. That’s fundamentally different than strawmanning.

    Further, as explained above, I was operating under the premise that “gods are supernatural”. I now understand clearly that you want to use words contrary to their normal usage, and I can work with that.

    One thing about the courtroom analogy again. If you listen to law lectures they really do talk about how “innocent” and “guilty” are not the same thing. I have read many a lawyer say when a person tells them they are innocent the lawyer respond with, “I am not interested in why you are innocent, I am interested in how you are not guilty.” I have seen a lecture (for the life of me I can;t find it but there are other boards where people talk about this) where innocence and guilt are drawn as Venn diagrams. The innocence and guilt circles of a particular crime (each crime is shown as a separate diagram) are disjoint (showing that one cannot be both innocent and guilty of the same crime) but the areas outside of each circle show that the region of not guilty includes the circle of innocence, but also includes a region of non innocence and non innocence includes guilt and non guilt. So non guilt and non innocence have overlap. I figured that you had heard this or WOULD look it up (I will give a link at some point).

    You’re just reinforcing my belief that you don’t know what you’re talking about. Again, please try to listen.

    You’re confusing events in the event space with beliefs in the belief space. In those lectures, the lawyers will also explain that the facts of the matter are that a defendant is either innocent or guilty (corwyn’s esoteric points are not relevant), but we have to operate on limited information. Because of our limited information, our beliefs about the facts are not binary “innocent or guilty”. Instead, our beliefs lie on a continuous spectrum from “convinced of innocence” to “leaning innocent” to “undecided / agnostic” to “leaning guilty” to “convinced of guilty” and everything in between. The “not guilty” verdict is a description of the belief-space. The “not guilty” verdict includes “convinced of innocence”, “leaning innocent”, “undecided / agnostic”, and “leaning guilty”. However, the facts of the real world are that the set of events which fulfill the boolean predicate “not guilty” is the set {innocent}.

    In other words:

    Event space: {innocent, guilty}. Exactly size 2 (give or take ambiguities in the law itself).

    Belief space: {convinced of innocence, leaning innocent, undecided / agnostic, leaning guilty, convinced guily}, and everything in between. The belief space is an infinite set. It directly corresponds to the Reals in the range [0, 1], which is the same range [0%, 100%].

    Again, we never have complete access to the event space. Instead, we have access to evidence which allows us to make probabilitistic e.g. confidence-level-based assertions about the event space. Formally, we don’t say “It is true that the event is guilty”. Instead, formally, we say “I believe that the event is guilty to a 99% degree of confidence”. We give a “not guilty” verdict when our degree of confidence of guilty is less than the “reasonable doubts” threshold, which I can describe as 99% for the sake of argument.

    Again, I think you don’t know the first thing about Bayesian reasoning, and you might try to listen to me describe it, and/or pick up the book Proving History by Richard Carrier who probably does a much better job than I in explaining Bayesian reasoning.

    The other things that the lecture gets at is how innocence and guilt of one crime does not equate to innocence and guilt of another. They also talk about innocence and guilt being highly conditional (so there are lots of “scenarios,” as you mentioned before, I just didn’t get into it). What I got out of it was something to the effect of P(Innocent) = 100% – P(guilty) – ( P(~innocent) and P(~guilty) ). That is oversimplifying it as guilt and innocence are much more complicated principles that are highly conditional, I am just using it to demonstrate that as far as the court is concerned, there is overlap between non guilt and non innocence (even of a single crime). It is probably because the question of whether what happened was a crime as well as whether the action was committed by the person in question. I think I get what you are getting at though, there is a certain range of sample space.

    Yep. You don’t know what you’re talking about w.r.t. Bayesian reasoning.

    Again, what you are doing is you are putting entities from the belief space in the P(). That’s wrong. That’s not how Bayesian reasoning works. You put the actual concrete events in the P(). The things inside the P() are things from the event space, and the resulting probability numbers describe the belief space. You should think of it like this. P is a function that maps from 1- the set of facts about the real world, to 2- the confidence level of your beliefs.
    P : facts -> beliefs.
    P : real world -> confidence levels.

    I agree that there is overlap of the “not guilty” verdict and a hypothetical “not innocent” verdict. A “not guilty” verdict encompasses all confidence levels of guilt from 0% to 99%, and a hypothetical “not innocent” verdict encompasses all confidence levels of guilt from 1% to 100%. The overlap is the range of confidence levels from 1% to 99%.

    Never put “not guilty” verdict inside the P(). That makes no sense (in this context). That’s wrong. “Not guilty” verdict is a description of your beliefs and your confidence levels. It is not a fact about the world, which is what goes in the P(). The truth of the matter of guilt or innocence, that’s a fact that goes in the P(). Our beliefs and confidence about the matter of guilt or innocence, that’s a confidence level which is described by the output of the P() and Bayes equation.

  217. frankgturner says

    @ EL
    I wasn’t so much mad at you as I needed to get that off my chest. I’ve reading your stuff for a while and you reminded me of a previous poster who was so self righteous and unwilling to consider how others think not give them the benefit of the doubt.
    .
    Fuzzy words are like that. People use them in variable ways. I recognize that and it’s with I ask a lot. I hear people use the word “theory” when they clearly mean hypothesis and I ask a lot. I here people say “truth” when they are clearly not talking about something that is “factually correct.”
    .
    Regardless of general consensus we don’t know for certain what people mean in exact terms. Its not like we are in the Matrix and can load their thoughts directly into our heads. I’ve known a LOT of people who express what I view to be too much certainty about context and intent.
    .
    I know that I need to work on my communication skills. We all do, it’s an ongoing process. So do you. The difference is I give people the benefit of the doubt a lot, sometimes even if I am VERY certain that I DO understand the context in which they are speaking. That is why I couch my terms a lot. I grew up with individuals who insisted that they understood the context in no uncertain terms and that they were absolutely right about how they read people.
    .
    That’s the thing about communication, it is a give and take process (metaphorically speaking). People who generally tend to insist that they are right about the context In no uncertain terms and often don’t give the benefit of the doubt are takers who don’t give (in my personal experience). Even without people specifically stating that they are speaking hypothetically it occurs to me that they might be. Maybe I give too much and don’t take enough.

    The metaphor about giving and taking when it comes to communication is that you don’t get to improve on your own terms. You make improvements in other people’s terms, otherwise you are just taking and not giving
    .
    I tend to ignore people like that eventually as I see them as bullies (as do many others). That is what happened to the former poster whom you remind me of (come to think of it, he may not have realized that people were ignoring him so it may be you after all).
    .
    I was the last hold out.
    .
    @ Narf or corwyn
    If you are listening, do you remember who I am talking about?

  218. EnlightenmentLiberal says

    Even without people specifically stating that they are speaking hypothetically it occurs to me that they might be.

    As a general rule, I think that’s incredibly ridiculous. If you tell me what you had for lunch, or how your day went at work, etc., I’m not going to ask you every time “Do you really mean that? Or was that a hypothetical for the sake of argument?”. I suggest you stop trying to weasel on this one.

    I was the last hold out.

    Pretty sure you’re thinking of someone else. IIRC, I had this name for long before you starting posting to these forums.

    This is one of your earlier posts back when you were posting under “Frank G Turner.”
    http://freethoughtblogs.com/axp/2014/07/03/determining-the-attributes-and-effects-of-gods/#comment-230954
    July 12, 2014

    This is one of the earliest posts I can find on this blog under my current name. I do not recall if I started posting here under my earlier name “codemonkey”. The post was under the name “EnlightenmentLiberal – formerly codemonkey”.
    http://freethoughtblogs.com/aronra/2013/04/20/inappropriate-invocations-of-the-rowlett-city-council/#comment-17107
    April 27, 2013

    I’ve been posting here for about a year before you came, and probably a bit longer under my older name. I don’t care enough to look.

    Also, it seems that you have thus declared that you’re going to ignore me. If you cannot handle when I tell you that you in no uncertain terms that you are wrong, and why you are wrong, and how you are wrong, but when I’m also open to counter-argument, and not because of excessively vulgarity, namecalling, etc., and you’re going to ignore me for that, then nothing of value of lost. I think your old religious ways are clouding your beliefs on this matter. You need to let go of respect for beliefs. It’s still clouding your thinking.

  219. Narf says

    @fgt – 231

    @ Narf or corwyn
    If you are listening, do you remember who I am talking about?

    Are you thinking of Adamah, or whatever his name on here was? It was Adam, with two other letters on the end, at any rate, if that’s the guy you mean.

  220. EnlightenmentLiberal says

    @Narf
    Wasn’t I also in several conversations with Adamah? I don’t understand why Frank thinks I’ve changed names in the time he has been here. I’m pretty sure that I’ve been here much longer than him with this name, and my google-fu seems to have confirmed that (see links above). Do you have any ideas? I’m completely lost on that one.

  221. Narf says

    *shrug*
    No idea what Frank means exactly. I’ve only been vaguely skimming this conversation. I’ve been very busy most of this week. I keep meaning to read through the comments on the post for Episode 919, but I haven’t yet.

    It just seems like that’s who Frank was referring to, based upon the paragraph or two that I read of #231.

  222. EnlightenmentLiberal says

    @Narf
    Sorry for the length. I would appreciate your input on my performance after you get around to reading it. I expect that I was harsh, perhaps needlessly so, but IMHO no worse.

  223. Narf says

    I’ll take a look at it, sometime in the next couple of days.  I’m running around in circles for most of today, as I have been for most of the past week.

    Was this conversation one that started with Bayes Theorem?  I haven’t looked back all the way, just scanned the last half dozen or so.

    I tend to dismiss Bayes’ Theorem in all but the most hard-scientific contexts.  I don’t see how it has any application to the Historicity of Jesus, for example.  When our documentation for a proposition is as shit as everything in Christianity is, and we have almost nothing that isn’t documentary evidence, the assignment of confidence to any particular element seems absolutely arbitrary.  I’ve seen shitty apologists using Bayes’ Theorem or something like it to conclude almost absolute certainty for the existence of their god.

    It has a massive GIGO failure, in almost every usage I’ve seen.  More to the point, the people in the aisles can’t follow it, so it’s useless when talking to them, anyway, unless you’re throwing shit against the wall to wow them with how brilliant you are, while you’re constructing an argument from authority.  It fails at several points, when I evaluate it to determine if it’s a useful tool to use when talking to random people.  You can bring people up to speed on syllogistic logic fairly quickly, so that’s a much better tool.

    In regards to the supernatural = nonexistence issue, I suspect that I’m a bit closer to your side, judging from the tiny bit that I’ve scanned.  But you need to coach it more in terms of functionality and usage, rather than the actual.  Believers resort to speaking about the supernatural after they’ve been beaten back from any hope of having an evidentiary argument for their position.  That makes the subject of their claims functionally nonexistent, since they’re basically making things up or accepting something that someone else made up, at that point.

    I dunno. I’ll go over it in greater detail and see if I have any more useful input, when I have the time.

  224. EnlightenmentLiberal says

    @Narf
    About Bayesian w.r.t. GIGO. Completely agreed. Still, I would urge you to consider this point: That’s precisely true to the same extent of forming regular syllogistic arguments. If the premises are bullshit, then it tells you nothing about the conclusion, and we’ve both seen Christian apologists do plenty of GIGO syllogistic arguments. I think the value of sometimes creating formal syllogistic arguments and formal Bayesian arguments is that between two honest people, it allows them to narrow down any points of disagreement.

    I’m much more concerned here that people learn some of the principles of Bayesian reasoning rather than the formula itself. Specifically, I want to do away with this silly notion that one can examine a claim in a vacuum without equal consideration of all plausible opposing claims.

    Also, thanks.

  225. Narf says

    Sure, you can cook the premises, in any form of logic.  Bayesian logic seems to lend itself more easily to obfuscation of that sort of activity, though, from what I’ve seen.  The further away that you take something from simple language, the less accessible it is.  That’s always my primary concern, when dealing with this stuff: simplification and accessibility.

    … is that between two honest people, it allows them to …

    And there’s precisely the issue, isn’t it?  When we’re dealing with the arguments of Christian apologists, we aren’t dealing with the arguments of honest people.  When the sheep come to us, waving around the arguments of the professional apologists, it’s easier to help them understand where the flaws are, if we stick with common language. Pointing out the equivocation fallacy in premise 1 of Kalam is far more effective than any amount of logical formulae.

  226. Monocle Smile says

    I’m much more concerned here that people learn some of the principles of Bayesian reasoning rather than the formula itself. Specifically, I want to do away with this silly notion that one can examine a claim in a vacuum without equal consideration of all plausible opposing claims

    Pretty sure that was covered really early in this thread, and you don’t actually need to consider “all possible” opposing claims, because there could be lots of unknowns. You just have to consider the other half of the true dichotomy.

  227. Monocle Smile says

    To elaborate, I’d like to point out that in the case of the proposition “the defendant is guilty,” in Bayesian reasoning, you need to consider the proposition “the defendant is not guilty” instead of “the defendant is innocent.” In my mind, there are important distinctions here.

  228. EnlightenmentLiberal says

    @MS

    Pretty sure that was covered really early in this thread, and you don’t actually need to consider “all possible” opposing claims, because there could be lots of unknowns. You just have to consider the other half of the true dichotomy.

    To elaborate, I’d like to point out that in the case of the proposition “the defendant is guilty,” in Bayesian reasoning, you need to consider the proposition “the defendant is not guilty” instead of “the defendant is innocent.” In my mind, there are important distinctions here.

    I tried to explain how that is an error of reasoning. In fact, that was more or less my whole goal of this conversation. I simply failed at explaining. Unfortunate.

  229. Monocle Smile says

    Or it’s not an error. I should have been more thorough…there are issues when multiple hypotheses would generate the same data set. Said data set would not be considered “evidence” for any of them, although it may reduce the possibilities.

    I understand what you’re saying, and I mostly agree. I’m just generally with Narf on this subject…it’s best to stick to language and concepts that are accessible unless it’s entirely necessary to do otherwise.

  230. Kudlak says

    @Narf #221
    There’s a bit going around these days where they say that even the devil can quote scripture, which excuses all past incidences where the Bible was used to hurt people. If you read it carefully, they say, you will uncover the overlapping theme of universal love which made freeing the slaves possible, for example. It also doesn’t hurt to blame all the past mistakes in theology on the Catholic Church, something that evangelicals love to use while poaching Catholics as new members of their churches.

    It seems ironic that these same people love to use quotes from scientists a 100 years ago and the argument that science is always changing it’s mind when it’s clear that they consider present theology to always be the correct one, and would call it unfair to quote some past theologian who supported slavery, for incidence.

  231. John Iacoletti says

    To elaborate, I’d like to point out that in the case of the proposition “the defendant is guilty,” in Bayesian reasoning, you need to consider the proposition “the defendant is not guilty” instead of “the defendant is innocent.” In my mind, there are important distinctions here.

    Agreed. Similarly the statements:

    “I believe some god exists”
    “I don’t believe some god exists”

    NOT

    “I believe no god exists”

  232. Narf says

    @246 – Kudlak
    Yup, the no-true-Christian fallacy is nothing new. They just used to call them heretics or something along those lines, up until a couple hundred years ago, and they used to just burn them at the stake or something.

    Yeah, universal love … I’d love to see an actual scriptural quotation on that one. The fuzzy theists who push this line never seem to know shit about what’s actually in the Bible.

  233. Kudlak says

    @Narf #248
    There’s every indication that there were competing brands of Christianity from the very beginning. Not only was Paul’s variety drastically different from that of Peter and James, I actually think that it was their form of Christianity that Paul was warning his followers against, not some weird Gnostic sect, which is commonly assumed by Christians. Wouldn’t that put the entire early Church into a different perspective if it could actually be proven true?

    Universal love is an impossibility within most forms of Christianity as they will always see themselves as the “wheat” which must be separated from the chaff, which happens to be everyone who doesn’t agree with them 100%. There’s also the whole persecution complex for them to justify, and you cannot claim to be a victim if you don’t garner criticism by excluding people, now can you?

    No, the appeal of the faith for many Christians is it’s tribal identity of holding the perceived truth, while simultaneously being part of a majority while being persecuted all at the same time. To them, it just wouldn’t be Christianity without having enemies, and any group claiming to be accepting of everyone and Christian just can’t be following their Bible, or Christ.

  234. corwyn says

    @244:

    I tried to explain how that is an error of reasoning. In fact, that was more or less my whole goal of this conversation. I simply failed at explaining. Unfortunate.

    Perhaps you failed because you are wrong?

    Take A = “guilty” then Bayes theorem gives
    P(B|A) * P(A)
    P(A|B) = ————————–
    P(B|A)*P(A) + P(B|~A)*P(A)
    It is easy to see from this that what we need to not is based on ~A, that is “not guilty”

    If you want to bring innocent into the picture you need to use a different formulation of Bayes’s theorem namely:
    P(B|A[i]) * P(A[i])
    P(A|B) = ——————————–
    SUM over j (P(B|A[j]*P(A[j]))

    where A[i] is ‘guilty’ and A[j] are all the possible situations, i.e. innocent, no criminal intent, insanity, etc.

    At that point, it is a matter of determining and *everyone agreeing* that all possibilities are being considered. If there is an argument, it can be solved merely by adding any possibilities under contention, as this can not affect the outcome if probability are assigned correctly. That you would still be arguing this, means that you have missed this fundamental point. Accept that other people seem to think that {guilty, innocent} is not a true dichotomy, allow them to add whatever they want to that set, and then perform the math.

  235. corwyn says

    wow total formatting fail.
    let me know if it is understandable as is.

    also: “…we need to not is based…” read “…we need to know is base…”

  236. Narf says

    @249 – Kudlak

    There’s every indication that there were competing brands of Christianity from the very beginning.

    That’s one of the reasons I don’t give much credence to the Christ-myth argument. Who was it pushing it on here, a week or so ago? Wasn’t that Hippycow?

    From what he was saying, it rests upon the contention that Christianity started as a mystery religion, and it mutated from that. I don’t buy that premise, so the whole argument is dead right out of the gate. Early Christianity was such a mess, and the Orthodox church was so busy with retcons that I don’t think anyone can build a reasonable case for any particular scenario. I’m sure that there were mystery-religion versions of Christianity, early on, but I won’t accept any argument that’s dependent upon any hard statement like that.

    Do I accept that Jesus actually existed? No.
    Do I accept the argument that the Jesus story is not based upon a real person? No.

    It might occasionally make for a fun speculative exercise, but I don’t consider it an argument worth having, if you’re expecting anything productive to come of it. And since I’m mostly interested in theistic arguments as they relate to the theists, I don’t care enough about this one.

    When addressing Christians, I find it best to move from the position of, “Look, we don’t know. There are many parts of the gospels that we know are flat-out wrong, and the rest is so unreliable as to be ridiculous. It’s so bad that many find it reasonable to argue that Jesus never existed at all, and they make an interesting case.”

    You can make use of the argument without actually arguing for it, which is pointless.

    Not only was Paul’s variety drastically different from that of Peter and James, I actually think that it was their form of Christianity that Paul was warning his followers against, not some weird Gnostic sect, which is commonly assumed by Christians. Wouldn’t that put the entire early Church into a different perspective if it could actually be proven true?

    I dunno, I’ve heard versions of this sort of thing. The most interesting evidence we have of the other early versions of Christianity is the writings of the Orthodox leaders, and those writings presented a badly distorted version of those groups, I’m sure, since those groups were competing versions of the same religion.

    I imagine it’s sort of like how Catholics and Protestants demonized each other far worse than they demonized Muslims, up until fairly recently. Hindus live in that almost-neutral mental-space, since there haven’t been any major assaults on our culture by them. Sure, if you asked any born-again Christians, I’m sure they would say that Hindus are going to hell, but you won’t get the same level of vitriol.

    Universal love is an impossibility within most forms of Christianity as they will always see themselves as the “wheat” which must be separated from the chaff, which happens to be everyone who doesn’t agree with them 100%. There’s also the whole persecution complex for them to justify, and you cannot claim to be a victim if you don’t garner criticism by excluding people, now can you?

    No, the appeal of the faith for many Christians is it’s tribal identity of holding the perceived truth, while simultaneously being part of a majority while being persecuted all at the same time. To them, it just wouldn’t be Christianity without having enemies, and any group claiming to be accepting of everyone and Christian just can’t be following their Bible, or Christ.

    That could be one of the big mental differences between fundamentalist and liberal Christians, yeah. Tribalism. Hmm.

  237. EnlightenmentLiberal says

    @corwyn
    Your list:

    where A[i] is ‘guilty’ and A[j] are all the possible situations, i.e. innocent, no criminal intent, insanity, etc.

    It’s all good points. It may be inaccurate to describe the event space as simply { innocent, guilty }. I may quibble a little along the way, but we’re not even that far yet with Frank and John. You’re at step 5 or 10, and they’re still stuck back at step 2. Their complaints and misunderstandings have nothing to do with your reasonable point here.

    PS: My quibbling would amount to this:

    “Not guilty by reason of criminal insanity” is a subset of “not guilty event” e.g. “innocent event”. It’s not a third option. Generally to be guilty of a crime that requires criminal intent, the law requires that you are sufficiently competent to have criminal intent.

    “No criminal intent” can be broken down to two smaller sets, one of which is a subset of “guilty” and one which is a subset of “innocent”. Some laws demand criminal intent for conviction, and others do not (often called statutory laws). “No criminal intent” is not a third option.

    Still, I appreciate this discussion. Still, I wish you would help me correct the fundamental misunderstandings of Frank, John, etc., before we move on to this advanced-by-comparison discussion.

    @Narf

    That’s one of the reasons I don’t give much credence to the Christ-myth argument. Who was it pushing it on here, a week or so ago? Wasn’t that Hippycow?

    I was somewhere recently. And I still think you would do well to read Richard Carrier’s “On The Historicity Of Jesus”, but I agree that it’s an esoteric point that doesn’t matter when talking with Christians due to 1- lack of mainstream scholarly acceptance, and 2- better defeater arguments against Christianity.

    The most interesting evidence we have of the other early versions of Christianity is the writings of the Orthodox leaders, and those writings presented a badly distorted version of those groups, I’m sure, since those groups were competing versions of the same religion.

    What writings? What leaders? We don’t have any of that. I hope you’re not talking about the gospels and acts.

  238. Narf says

    I’ll read his stuff, at some point.  He just has a lot of work to do before I’ll take it seriously, and I don’t think that work can even be done.

    What writings? What leaders? We don’t have any of that. I hope you’re not talking about the gospels and acts.

    Oh, fuck no.  😀  I doubt any actual history could be mined from those documents with anything approaching an acceptable level of confidence.

    I’m talking about stuff later than that.  The doctrine wars between the Orthodox, Ebionites, Gnostics, Docetists, and other various groups.  There was so much forgery and slander going on that I don’t think we can pick it apart, this far removed from the events, particularly after the Orthodox paved everything over.

  239. Narf says

    I think that another way you could look at it is comparing Jesus to Socrates, only with an additional order of magnitude or two of dubiety thrown in.

  240. EnlightenmentLiberal says

    @Narf
    Sure, but IIRC there’s no way we can identify if those later writings came from the Paul sect or some competing sect. IIRC, all we can say is something like the Paul side won, because the gentile Christianity triumphed over the Jewish observant Christianity. Beyond that, I think we have basically nothing. All of the writings we have basically give zero information aobut any sects competing against Paul in Paul’s time.

  241. Narf says

    That’s essentially what I said a few comments back, yes.  “The most interesting evidence we have of the other early versions of Christianity is the writings of the Orthodox leaders, and those writings presented a badly distorted version of those groups, I’m sure, since those groups were competing versions of the same religion.”

  242. Kudlak says

    @Narf #252
    Yeah, if it was the tread about Carrier’s new book I think I was one of the ones really pushing against the theory, but that may be mostly because of my past experience with attempts to demonstrate the idea, all of which were pretty lame. If I remember correctly, one of the key points that Carrier makes is that the very earliest Christians all seem to be venerating a spirit, with the Gospels only later adding details about a human life for Jesus.

    To me, this only makes sense, as the “earliest Christians” here where converts of Paul, who had only ever claimed to know the risen Jesus through revelation, owing to Jesus’ crucifixion being some years earlier. Being in competition with the “super apostles” who had actually known the living Jesus, and claiming to get his own, unique Gospel directly from the risen Jesus, Paul of course would have deemphasized the life of his Lord and focused upon what Jesus could do for people now, as the spirit Christ.

    That still leaves open the possibility that a risen Jesus was entirely Paul’s creation, perhaps even a delusion conjured up by his mind in order to place himself on the same (or superior) footing as the apostles. If those guys actually taught against the notion of a risen Jesus who could be around to endorse Paul it would be little wonder why he warned his followers against listening to them, right?

    Do I accept that Jesus actually existed? Jesus, maybe, but Christ? No. Christ (at least an active spirit of Jesus able to contact people before his second coming) seems more likely to be an invention of Paul’s in order to give him a means to claim apostleship.
    Do I accept the argument that the Jesus story is not based upon a real person? Like you, no.

    To me, it’s like the Zeitgeist material, and it only serves to give believers a rather easy excuse to dismiss all arguments against their literal interpretation of the resurrection.

    Paul is very clear about the difference between his gospel and that of the super apostles James and Peter. Although Acts glosses it over, even it says that several meeting were necessary before both sides basically agreed to disagree.

    Catholics and Protestants are also competing for the same pool of potential converts, so it’s more like the battle for cola drinkers with Coke vs Pepsi, with little attention given to the Sprite, Mountain Dew and 7-UP drinkers out there who are less likely to make the switch. 🙂

  243. corwyn says

    @ EL 253:

    a subset of “not guilty event” e.g. “innocent event”.

    This is exactly where everyone disagrees with you. By fighting this battle instead of the one you claim to care more about, with John, you are losing *that* battle. There is no reason to fight it when it distracts from your goal. A good Bayesian should feel free to include *any* wacky hypothesis in that sum confident that the truth will out.

  244. corwyn says

    @258:

    Being in competition with the “super apostles” who had actually known the living Jesus, and claiming to get his own, unique Gospel directly from the risen Jesus, Paul of course would have deemphasized the life of his Lord and focused upon what Jesus could do for people now, as the spirit Christ.

    I just don’t see how Paul could have won this argument, if he was arguing against people who had really talked with a living Jesus. Why would anyone give him the slightest attention?

  245. Narf says

    @258 – Kudlak

    To me, this only makes sense, as the “earliest Christians” here where converts of Paul, who had only ever claimed to know the risen Jesus through revelation, owing to Jesus’ crucifixion being some years earlier.

    Errrr, so who were the people who Paul claimed to have been persecuting, before Jesus smacked him around?

  246. John Iacoletti says

    @ EL 253:

    a subset of “not guilty event” e.g. “innocent event”.

    This is exactly where everyone disagrees with you.

    Exactly. It’s not a “fundamental misunderstanding”, EL, it’s a disagreement with your premise that “not guilty” and “innocent” are the same thing.

  247. Narf says

    @260 – corwyn

    I just don’t see how Paul could have won this argument, if he was arguing against people who had really talked with a living Jesus. Why would anyone give him the slightest attention?

    I see three big problems there.

    For one thing, why would anyone pay him any attention?
    A little charisma goes a long way, when gathering cult followers. Even if the gospels bore any resemblance to history … which I have no reason to accept … people who were in the original political movement of the historical Jesus (if such a person existed) wouldn’t necessarily have more success than someone who did a better job of making up an appealing story and selling it to the converts. I don’t know that Paul was all that charismatic, but if he was, that could explain pretty much everything.

    Second, who says that Paul actually argued against any of the actual followers of an historical Jesus? The only things we have to go on for that assumption are the Gospels and Acts, which are so far from reliable as to be silly. Perhaps all of the first-hand witnesses to the historical Jesus were already dead by the time Paul showed up, and Paul was arguing against the followers of the apostles. The whole narrative of Acts could have just been made up completely, after the fact, or pieced together from later stories and filled in with wishful thinking.

    We know that early Christians stuffed shit into the gospels, after the fact. With the earliest writings we have, we see that the passage about Jesus and the woman taken in adultery was inserted, as well as the agony in the Garden of Gethsemane. Those are just some of the things we can detect that were added since our earliest versions.

    Who knows what was added, prior to that? You tend to get more volatility at the beginning of movements. If Christians were still engaging in that sort of behavior, after our earliest copies of the Gospels were made, how much more was it happening before then?

    Third issue … who says that Paul won the arguments? We only have Paul’s word to go on (or that of the people who forged letters in his name) that he won the arguments that he had in person. Even with the genuine letters, some of them bear the characteristics of someone who is trying to retroactively claim victory, after getting his ass handed to him. Paul’s writings survive, which is a victory of sorts, but without the opinions of the people who watched the debates in person, we can’t assume anything.

    I’m sure that if you ask William Lane Craig, he’d say that he’s won every debate he’s ever had, and his opponents were buffoons who don’t understand the first thing about philosophy … and that philosophy is subordinate to the self-authenticating witness of the holy spirit, anyway, and since none of his opponents brought along the holy spirit on their side …

    Hell, Shock of God, on YouTube, explicitly says that. He’s apparently won every debate he’s had, if you listen to him, because the atheists failed to provide proof, to his satisfaction, that God doesn’t exist.

  248. EnlightenmentLiberal says

    @corwyn
    You don’t understand what’s going on. Again, your complaint is not the problem with Frank and John (and Matt, etc.).

    It’s really funny that John thinks he’s agreeing with corwyn, but they’re making completely differnet points.

    It’s said that a good debater and communicator can make the same point of their opponent better than their opponent. Let me try that now.

    Matt, John, and others from the show have this meme. They argue that the only obviously true dichotomy is “X” and “not X”. Let me try to unpack exactly what they mean. Take a really simple claim. “Alice stabbed Bob 47 times with a knife (in a span of a few minutes).” Suppose we present evidence that Bob was stabbed 47 times with a knife in a span of a few minutes in his house. Suppose we present evidence that Alice was driving her car near Bob’s house at the time of the stabbing. Matt, John, et al, would advocate thinking about this claim as “Alice stabbed Bob 47 times” or “Alice did not stab Bob 47 times” because that is the true dichotomy. If someone tried to frame the discussion in terms of “Alice stabbed Bob 47 times” or “Charlie stabbed Bob 47 times”, Matt, John, et al, would argue against by saying that these two new claims do not constitute a true dichotomy, and thus it’s fallacious. Further, they have often said the intellectual equivalent of “I do not need to consider the scenario where Charlie stabbed Bob and its associated evidence in order to consider the claim that Alice stabbed Bob and accept or reject the claim.”

    I say that the proper way to think about problems like this is to break the space of epistemically plausible / possible real world events, the event space, down into mutually exclusive events that cover the entire space (or at least down into events that are well defined with well defined overlaps). The idea that you can consider the claim “Alice stabbed Bob” without positively considering the claim “Charlie stabbed Bob” is wrong and outright foolish. Matt, John, et al would disagree, and I’m still at a loss how. I’ve tried to explain, but their replies indicate a near complete lack of understanding of everything at hand.

    Now you corwyn come in and complain, quite reasonably, that maybe Dana stabbed Bob. I’m ok with that. That’s a discussion worth having. However, it does not address at all the fundamental mistake where Matt, John, et al promote bad reasoning where they decide to accept or reject the truth of the claim “Alice stabbed Bob” without considering “Charlie stabbed Bob”, “Dana stabbed Bob”, and so forth.

    Again, in short, my position: Something counts as evidence for a proposition if and only if it’s more-expected on that proposition compared to all competing propositions. In order to determine if you should accept or reject a claim, you need to determine if there is sufficient evidence to accept the claim. In order to determine that, you need to positively consider all competing propositions and see if the evidence matches those propositions, because only by doing that can you determine if the evidence is actually in support of the proposition, or if the purported evidence is merely a non sequitir.

    To relate it to the position of Matt, John, et al, I would say that in all but the most trivial of examples, you cannot accurately determine how well the evidence fits the proposition “not X”. To use the example, if you think about the problem as “Alice did not stab Bob”, you have no idea how well the evidence fits that proposition. Only by breaking that down into concrete specific positive claims like “Charlie stabbed Bob”, “Dana stabbed Bob”, etc., can you even begin to start determining how well the evidence fits the competing propositions, and only by that process can you determine how much the evidence supports the initial claim, and only by that process can you intelligently determine to accept or reject the claim.

    There are some complications to this approach, but what approach doesn’t have its complications? This approach also comes with the added benefit of actually being correct.

  249. Kudlak says

    @Narf #261
    “Here” being the letters of Paul, where his Christians (rather than those from the James/Peter camp) are being addressed. Sorry if that wasn’t clear.

    His converts probably would not have been amongst those who were already followers when Paul was Saul, living far from Jerusalem many years after the crucifixion, right? So, when he talks about being a persecutor of Christians he means those earlier Christians from the other, original tradition who may have even known Jesus personally. He obviously disagreed with what they believed back then, and maybe his Damascus Road experience was nothing more than a sudden insight where he managed to square basic Christian beliefs to his way of thinking, fixing all the things about the original faith that made him oppose it. The irony then would be that he might have given up persecuting those Christians, but he never stopped opposing them.

  250. Kudlak says

    @corwyn #260
    Easy to win the argument when you are the one bringing the entire Christian message, including all information about Jesus, to people. Remember that his churches were hundreds of miles outside of the land of the Jews, in Asia Minor and other parts of the Empire, and that the Epistle to the Hebrews is now considered by most scholars to be a forgery in Paul’s name. It was his idea to bring Christianity to these gentile areas, and his churches could have been operating for many years before making significant contact with folks from the original Jewish brand of the faith. That would have caused exactly the kind of confusion that Paul addresses in some of his letters.

  251. Monocle Smile says

    @EL

    Suppose we present evidence that Alice was driving her car near Bob’s house at the time of the stabbing. Matt, John, et al, would advocate thinking about this claim as “Alice stabbed Bob 47 times” or “Alice did not stab Bob 47 times” because that is the true dichotomy. If someone tried to frame the discussion in terms of “Alice stabbed Bob 47 times” or “Charlie stabbed Bob 47 times”, Matt, John, et al, would argue against by saying that these two new claims do not constitute a true dichotomy, and thus it’s fallacious

    Along the right lines somewhat, but that’s extreme.

    Further, they have often said the intellectual equivalent of “I do not need to consider the scenario where Charlie stabbed Bob and its associated evidence in order to consider the claim that Alice stabbed Bob and accept or reject the claim.”

    I agree with this entirely.

    I say that the proper way to think about problems like this is to break the space of epistemically plausible / possible real world events, the event space, down into mutually exclusive events that cover the entire space (or at least down into events that are well defined with well defined overlaps). The idea that you can consider the claim “Alice stabbed Bob” without positively considering the claim “Charlie stabbed Bob” is wrong and outright foolish. Matt, John, et al would disagree, and I’m still at a loss how. I’ve tried to explain, but their replies indicate a near complete lack of understanding of everything at hand

    Okay, this is much clearer. I think I see the problem: I believe you’re just flatly wrong. It’s not at all necessary to do what you claim in this situation. It may be useful, and may swing the pendulum further in some direction and get us closer to figuring out the actual state of affairs, but it’s not remotely necessary if we’re only talking about the boundary between “belief” and “no belief.”

  252. Monocle Smile says

    To use the example, if you think about the problem as “Alice did not stab Bob”, you have no idea how well the evidence fits that proposition. Only by breaking that down into concrete specific positive claims like “Charlie stabbed Bob”, “Dana stabbed Bob”, etc

    Again, this is wrong, because absence of evidence is evidence of absence, and evidence like Alice having an alibi does indeed fit that proposition much better than “Alice did stab Bob.”

  253. EnlightenmentLiberal says

    Sorry, I should go on.

    Or video tape showing Charlie is the one who stabbed Bob.
    Or phone recordings of Charlie in a conversation to hire a hit man to stab Bob.
    Or strong motive like a large amount of money for Charlie if Bob died.
    And so on.

    It’s trivial to imagine evidence specific to Charlie which should change your estimation of the likelihood of the truth that Alice stabbed Bob, and only by also taking seriously the scenario of Charlie stabbing Bob (or ordering it – another scenario) can you accurately determine the likelihood of the truth of Alice stabbing Bob.

  254. Monocle Smile says

    only by also taking seriously the scenario of Charlie stabbing Bob (or ordering it – another scenario) can you accurately determine the likelihood of the truth of Alice stabbing Bob

    I disagree entirely. That is merely one method of doing so. Did you not read my post #267? If I’m sitting with Alice when Bob was stabbed, then I don’t need to consider the scenario of Charlie stabbing Bob, do I? Not unless we’re going to toss Occam’s Razor out the window and go with the cyborg doppelganger theory.

  255. bkilgore says

    A bit off topic but I hope the hosts get this and possibly use it in one of their episodes but has anyone actually thought about these points when thinking of the rules of the bible:

    1. Jealousy is a sin.
    2. Anyone who sins, or who follows sinners, goes to hell.
    3. God admitted he is jealous thus inadvertently an admitted sinner.
    4. Going by the rules and following the logical deduction . . . IF YOU FOLLOW GOD YOU GO TO HELL.
    5. It is in the bible so try to prove that wrong.
    6. To prove it wrong would be to disprove the bible.
    7.To prove the bible wrong with the only intention to prove these points wrong just so you could be right would make you against your own religion, a.k.a. an atheist.
    8. Bazinga! Faith just got butt-raped by its own rules.

    Anywho just wondering if anyone else ever noticed this nice little loophole?

  256. corwyn says

    @EL:

    Only by breaking that down into concrete specific positive claims like “Charlie stabbed Bob”, “Dana stabbed Bob”, etc., can you even begin to start determining how well the evidence fits the competing propositions

    Can you imagine the courtroom scene? “…but your honor, we are only at ‘BA’ in the phone book, we need to consider everyone before we can decide on the defendant’s guilt.”

    If there is live coverage of Alice at a basketball game during the time of the stabbing, that is likely sufficient evidence without considering anyone else as a culprit. Yes, implicit in that is the idea that surely someone else must have done it instead, that need not be explicit for a trial.

    All we really need is an accurate denominator in Bayes’ equation (i.e. A & ~A, or A[1-n] where Sum p(A[i]) = 1).

  257. John Iacoletti says

    @EL 264

    To relate it to the position of Matt, John, et al, I would say that in all but the most trivial of examples, you cannot accurately determine how well the evidence fits the proposition “not X”. To use the example, if you think about the problem as “Alice did not stab Bob”, you have no idea how well the evidence fits that proposition. Only by breaking that down into concrete specific positive claims like “Charlie stabbed Bob”, “Dana stabbed Bob”, etc., can you even begin to start determining how well the evidence fits the competing propositions, and only by that process can you determine how much the evidence supports the initial claim, and only by that process can you intelligently determine to accept or reject the claim.

    There are some complications to this approach, but what approach doesn’t have its complications? This approach also comes with the added benefit of actually being correct.

    There’s that arrogance again. You think that it’s “the proper way to think about problems like this”, therefore it’s “actually correct”. This is only a problem for you because you continue to confuse “I don’t believe Alice stabbed Bob” with “I believe Alice didn’t stab Bob”.

  258. EnlightenmentLiberal says

    @John
    Let me try this pithy phrasing.

    Any evidence that implicates Charlie necessarily exonerates Alice. Do you agree? Any evidence circumstantially in favor of Charlie’s guilt necessarily is circumstantially in favor of Alice’s innocence. Do you agree? (Ignoring scenarios where they did it together, which is also important to take into account, but I’m trying to make a point here.)

    Any evidence that exonerates Charlie necessary raises our estimation of Alice’s guilt. Not to the same degree. It depends on the population size e.g. the number of competing hypotheses, and it depends also on the other relevant evidence, but knowing that Charlie didn’t do it means someone else did, which must raise our estimation of Alice’s guilt – raising only slightly, but still raising. Do you agree?

    Further, suppose we also have really good evidence that restricts the number of possible guilty persons to a small number, like a dozen. Any evidence that exonerates Charlie necessarily raises our estimation of Alice’s guilt. Similar to before, the change in estimation of Charlie’s innocence is not the same as the change in estimation in Alice’s guilt. However, because the population size is now 12, the corresponding change to the estimation of Alice’s guilt is substantial – around 1/12 as big. Do you agree?

    Let’s go further. Suppose we have really good evidence that only Alice, Charlie, or Dana did it. With that evidence, we should start with a 1 out of 3 odds that any one of them is guilty. Suppose we add some good evidence that Alice didn’t do it. That means we should adjust our estimations to be something like 1 to 1000 odds that Alice did it, and 1 out of 2 odds that Charlie did it, and 1 out of 2 odds that Dana did it.

    Let’s go further. If we have good evidence that restricted it to two people Alice and Charlie, and if we also had good evidence that strongly exonerated Alice, then that incriminates Charlie, and a good jury should vote “guilty” for Charlie, despite the lack of incriminating evidence particular to Charlie beyond the starting evidence that either Alice or Charlie did it.

    Now, about your claims of arrogance. The difference between arrogance and other kinds of conviction is in the eye of the beholder. The difference between arrogance and other forms of conviction is that the beholder views one as correct, and the other as wrong. I politely but strongly disagree with your assessment. I again assert in strong terms that you are wrong, and if you wish to view that as arrogance, then so be it. This is me not caring. If you don’t want to engage, I will stop engaging, but I’m not going to back down off of my firm convictions just because you think I’m rude or arrogant or whatever other names you want to throw my way. This an ad-hom, or a baby step away from it.

    @MS
    Only because you have implicitly considered all of the other evidence, and have implicitly determined that there is no equally compelling evidence for Charlie or Dana. However, I could introduce equally compelling evidence that Charlie did it. Admittingly, it’s getting rather extreme if I need this evidence to be convincing to you, and you personally witnessed the murder. However, that’s a rather extreme case. In most cases, jurors are not direct witnesses.

    @corwyn
    If the prosecutor has not presented evidence which fits Alice better than the average member of the population, then it’s not evidence in favor of the guilt of Alice. Your hypothetical statement to the judge and jury is oddly phrased, but still eminently correct. You of all people should realize this. You are the one who argued up-thread that our prior estimate of guilt before any evidence is presented should be 1 over the population size.

    Let me give a couple of examples with better phrasing:

    “We the defense have presented compelling evidence that the housekeeper probably did it, much more compelling than the evidence which the prosecutor has presented against Alice, which thus exonerates Alice. Please vote not guilty.”

    “We the defense observed that the prosecutor has presented no evidence which implicates Alice any more than a random member of the population. Yes Alice has the physical range of movement, the capability, of stabbing Bob, but so does Charlie and Dana, and many other people, thus that purported evidence does not favor guilt.”

  259. EnlightenmentLiberal says

    @MS
    Hell, you even explicitly considered some of the competing scenarios, such as robot impersonators, but you judged them too unlikely to be of note. Exactly like a Bayesian. You are already thinking like a Bayesian, explicitly so, but you don’t recognize that this is the form of your reasoning.

  260. frankgturner says

    @ John
    I have already gotten his point and it is a good one, but this isn’t about that anymore. It seems to be about an individual who “needs” to be “right” and needs everyone else to be “wrong” who disagrees. This is like listening to an apologist make the same tired argument over and over again (because their “God” MUST be the RIGHT one).
    .
    I dunno, would you ban an apologist from the message board after a while over this kind of insecure unprofessionalism?

  261. John Iacoletti says

    Any evidence that implicates Charlie necessarily exonerates Alice. Do you agree? Any evidence circumstantially in favor of Charlie’s guilt necessarily is circumstantially in favor of Alice’s innocence. Do you agree? (Ignoring scenarios where they did it together, which is also important to take into account, but I’m trying to make a point here.)

    You can’t ignore that though, even to make a point — it’s central to the entire argument. Evidence that implicates Charlie in and of itself does nothing to exonerate Alice.

    I’ll try being pithier. The only question in Alice’s trial is “did Alice kill Bob (beyond a reasonable doubt)”. It’s not “who killed Bob”.

  262. Monocle Smile says

    Hell, you even explicitly considered some of the competing scenarios, such as robot impersonators, but you judged them too unlikely to be of note. Exactly like a Bayesian. You are already thinking like a Bayesian, explicitly so, but you don’t recognize that this is the form of your reasoning

    Except I’m considering the true dichotomy, not a false one. Furthermore, I think you missed my point entirely. There are plenty of ways by which we can raise our “confidence gauge” past the “belief” line, and yet you seem to think there’s only one possible way to do this. There are important differences between possible, useful, and necessary.

    “We the defense observed that the prosecutor has presented no evidence which implicates Alice any more than a random member of the population. Yes Alice has the physical range of movement, the capability, of stabbing Bob, but so does Charlie and Dana, and many other people, thus that purported evidence does not favor guilt.”

    This actually goes against your point, except instead of stating that there’s no evidence that implicates Alice any more than a random member of the population, you just have to state that the presented evidence doesn’t fill your gauge past the “guilty” line. That’s it.

    Further, suppose we also have really good evidence that restricts the number of possible guilty persons to a small number, like a dozen. Any evidence that exonerates Charlie necessarily raises our estimation of Alice’s guilt

    Sure, but evidence that exonerates Alice doesn’t raise our estimation of anyone’s guilt relative to anyone else in that small number, and now it can be believed that Alice did not stab Bob even if we didn’t narrow down the number of people.

  263. EnlightenmentLiberal says

    @John

    You can’t ignore that though, even to make a point — it’s central to the entire argument. Evidence that implicates Charlie in and of itself does nothing to exonerate Alice.

    I want you to think about it.

    To make you think about it, I want to use a realistic example. Perhaps a real world example would be better, but I think this fictional example will be better because of your greater familiarity with it. Have you seen the great movie The Shawshank Redemption? I assume so.
    SPOILERS
    What follows may be spoilers. If you haven’t seen the movie, you really should.

    In the beginning of the film, the main character Andy Dufresne is (wrongly) convicted of murder. He goes to prison. Most of the film is set in prison. For most of the film, everyone in the prison was pretty well convinced that he was guilty of the murder.

    Later in the film, this happens:
    http://en.wikipedia.org/wiki/The_Shawshank_Redemption

    In 1965, Tommy Williams is incarcerated for burglary. He joins Andy’s and Red’s circle of friends, and Andy helps him pass his G.E.D. exam. In 1966, Tommy reveals to Red and Andy that an inmate at another prison claimed responsibility for the murders Andy was convicted of, implying Andy’s innocence.

    I don’t think I could say it better than Wikipedia.

    After this revelation – after this reveal this reveal of additional evidence that someone else did it, everyone seems to adjust their estimation of the likelihood of Andy’s guilt. Do you think Andy’s fellow prisoners were irrational in this conclusion? Would you not use the same reasoning to arrive at roughly the same conclusion if you were in their place? I think you would do the exact same (good) reasoning.

    Note: I am not saying that this flimsy evidence alone should be enough to convince you of Andy’s innocence. I am saying that in light of the new evidence, you should adjust your estimations of Andy’s guilt and innocence appropriately.

    This point about “Andy and the unnamed cellmate might have worked together” is largely besides the point. You are not going to complain about the reasoning of the prisoners in the movie because they didn’t consider this scenario. It’s a red herring. I can explain in detail why it’s a red herring in this case, and why it’s an important consideration in other cases, but first you need to recognize that what I quoted of you at the top of this post is wrong as a blanket assertion. You can salvage something of it, but in doing so you’ll basically end up right at my position.

  264. EnlightenmentLiberal says

    @MS

    Further, suppose we also have really good evidence that restricts the number of possible guilty persons to a small number, like a dozen. Any evidence that exonerates Charlie necessarily raises our estimation of Alice’s guilt

    Sure, but evidence that exonerates Alice doesn’t raise our estimation of anyone’s guilt relative to anyone else in that small number, and now it can be believed that Alice did not stab Bob even if we didn’t narrow down the number of people.

    This is just flagrantly irrational.

    Again, think like a betting person, like a gambler. Do you think the odds that a casino would offer on bets for the guilt or innocence of Charlie are going to remain unchanged after some strong evidence comes to light which exonerates Alice when the existing evidence has limited the plausible pool of suspects to less than 12? I think not.

    It would be foolish to deny the obvious truth of this hypothetical gambling establishment.

    It would also be foolish to deny the utility and correctness of adopting a gambler mindset like this for describing your confidence levels in your beliefs. Only by having certain levels of confidence in your beliefs can you do proper risk analysis and cost-benefit analysis, and only by doing proper risk analysis can you be a rational agent. Every day is a gamble. Every decision is a gamble. You never have absolute confidence. Every choice you make is on incomplete information.

  265. Monocle Smile says

    @EL

    After this revelation – after this reveal this reveal of additional evidence that someone else did it, everyone seems to adjust their estimation of the likelihood of Andy’s guilt. Do you think Andy’s fellow prisoners were irrational in this conclusion? Would you not use the same reasoning to arrive at roughly the same conclusion if you were in their place? I think you would do the exact same (good) reasoning

    When the shit has anyone in this thread argued otherwise? Are you not paying attention? No one is disputing this point. What we dispute is your apparent assertion that (in the previous analogy) that we MUST determine exactly who stabbed Bob or at least have evidence that a specific other person stabbed Bob in order to exonerate Alice. For the THIRD time, we’re talking about whether it is necessary to do so, not whether it would be useful or bring you to the same conclusion. There are multiple ways of reaching the same conclusion, which is exactly why defense councils sometimes offer alternate theories of the crime and sometimes don’t.

  266. Monocle Smile says

    Do you think the odds that a casino would offer on bets for the guilt or innocence of Charlie are going to remain unchanged after some strong evidence comes to light which exonerates Alice when the existing evidence has limited the plausible pool of suspects to less than 12? I think not

    Only if we’re talking about the population at large vs. those 12. The odds change, but each of the 12 still has the same odds relative to the other members of the 12. They all change the exact same amount if Alice is exonerated. Plus, we’re not talking about who stabbed Bob or if Charlie stabbed Bob. We’re only talking about whether or not Alice stabbed Bob. I think you totally misread my post.

  267. EnlightenmentLiberal says

    Ack. At myself in 281. I misread MS completely. My bad. Strike that post entirely.

  268. frankgturner says

    @ MS # 282
    I think that he is having imaginary conversations in his head about how people “would” respond, not how they actually “did” respond. Which should tell you something about his analysis of what he himself says. Sometimes he seems so insecure about his arguments that he is responding to criticisms that people have not even made but which he seems to think people are going to make. He seems to have responded to several things he thought that I or John or Matt, etc. were going to say but never actually did (or responded to me with regard to a position that Matt or someone else took).
    .
    My whole point about the implicit vs explicit consideration had to do with that while you might, to be intellectually honest, place the probability of someone else’s guilt into a mathematical equation of the calculation to calculate the probability of the guilt of the person being formally accused (in his case “Alice” whom I would point out is hypothetical but it is uneccessary at this point), you don’t necessarily make a formal statement about it unless the numbers are significant.
    .
    I don’t think this is about that any more so much as it is about an individual who has to be right on every single point. I have seen this before but it is typically apologists who do this (Adamah did as well though). It has definitely given me some insight into how apologists think.

  269. Kudlak says

    @bkilgore #272
    According to Christian apologetics, the rule against jealousy is one imposed by God upon humans, and not one that he applies to himself. It’s like the different rules you may impose in your house about eating at the table, where you may, but your pets may not, for instance. The average Christian will not even accept the idea that a mere human has the right, much less the ability, to judge God, so they are likely to just reject your whole argument out of hand.

    What you can ask your Christian friends is what’s so bad about jealousy? Sure, being obsessively jealous of somebody else probably isn’t healthy, but isn’t ambition and the basic drive to better oneself not linked to wanting something better, and how would you even know that something better is even possible if you did not see others with it?

  270. frankgturner says

    @ Kudlak # 286
    In other words “do as I say not as I do” from a father figure right?
    .
    (This hypothesis that I proposed of religion being an extension of primate evolution has a lot of points going for it).

  271. Narf says

    I don’t understand how Christians can’t look at this sort of thing and automatically call bullshit on the whole construction of their god concept. Their god is perfect … but he suffers from the same character flaws as an insecure tween girl. He’s a loving father, but he displays none of the attributes of such …

  272. frankgturner says

    @ Narf #288
    I have an idea based on a hypothetical proposal that fits the facts. That God they worship is really an extension of themselves or possibly an alpha male patriarch that they want to protect them. It can’t be wrong or falsifiable because then, in principle, it could be debunked. However, it has to be real on some level so an effort is made to argue (at least verbally) that it is falsifiable.
    .
    Of course we have seen that you need not be Xtian to project an image of oneself that is perfect and cannot make mistakes or be flawed, which is exactly whyvsaid being demonstrates insecurities. Its a catch 22 really.

  273. Kudlak says

    @frankgturner #287
    No, I actually do see it as more like my pet analogy, where the law-giving entity sees itself as a superior species with the right to dictate what’s best for us owing to it’s vastly superior nature. Male dominance within a species is more like our own power structures, excluding the most extreme forms of inherited autocracy, of course, where leaders like North Korea’s Kims are seen as gods for all intents and purposes. Type A personalities may seek out political office, but some type Bs find themselves elevated to positions of authority as well, and theoretically, anyone can become a leader, and every leader may be required to take leadership from somebody superior to them as well. What you’re describing is a struggle within a species to choose a leader. What I’m describing is a stronger species subjugating a weaker one, which I think best describes the God/man relationship.

  274. Narf says

    @289 – fgt
    Heh, yeah, I remember, from other times you’ve mentioned that. Sorry if I normally don’t have much to add on the subject. The extent of my thinking is usually something along the lines of, “Yeah, that sounds about right,” and I don’t have much constructive to add.

  275. frankgturner says

    @ Kudlak # 290
    Ok I think I see what you are getting at and that makes a lot of sense.
    .
    I would ask though if it could be e tended to the stronger group of individuals within a species subjugating a weaker group of individuals within that same species? That would go along with the idea of. devaluing and dehumanizing others whom you intend to be slaves which is pretty common in scripture. Your thoughts?

  276. frankgturner says

    @ Narf #291
    A lot of the wording for that comes from corwyn . I proposed ideas leading to it but he really combined them succinctly into a single statement.

  277. Narf says

    @290 – Kudlak
    Err, Frank is talking about the thought processes and motivations of the people who constructed the god-concept of the Christian deity, not the interaction of those people with their god, within the mythological world in which that god actually exists.

    Frank is trying to answer why people would imagine a being like the Jewish god as being above them and being grateful and subservient to such a monster. He’s saying that it goes back to simpler tribal structures, in which people would attempt to recreate the abusive, aggressive, alpha male who lead the pack, and the creation of the abusive gods was a reflex, trying to feel comfortable in an older, familiar setting of smaller tribalism, extended into the larger social structures that were developing.

    Basically, it’s the imagineering of an alpha male who could take up such a role over such a large societal unit.

  278. Kudlak says

    @ Narf #288
    God is like a shepherd, but not the kind who would ever fleece his sheep or slaughter them for meat (if such a thing ever existed back then.)

    God is like the wind, but not the nasty destructive kind that killed Aunt Bee when her trailer got tossed a quarter mile, and not the kind blowing smog and stinky garbage around either.

    God is like oxygen, but not the kind of oxygen that is easily combustive, or that you depend upon plants to provide.

    In the long list of things that God is like I cannot think of a single one that doesn’t also reflect one of his flaws. Maybe they should just stop trying to make analogies then?

  279. Narf says

    In other words, I think you and Frank are addressing two completely different scenarios, Kudlak.

  280. Narf says

    Err, are you talking about the god actually presented by the Bible or the kind that modern, fuzzy Christians insist actually is, despite the contradictory witness of the Bible?

    In all three of your statements, I identify the second part as being the god of the Old Testament. He slaughters his sheep for meat all the time, beats the Jews with the nations around them when they do the slightest wrong thing, and creates things like Satan, who burns Christians for being too good. Either I’m confused, or the rationalizing Christians are.

  281. Kudlak says

    @Narf #294
    I see what you’re talking about, but what I’m saying is that they started with the instinct to be led by a powerful member of their own species, and had to extend that to being led by a entity from another species that could, in their imaginations, exceed all the limitations that their own leaders had. You may be able to hide things from your parents, for instance, but not from Santa who has the ability to know when you’ve been bad or good. With gods we go that extra step by imagining the alpha male, but with powers beyond what we could ever suspect from someone from our own species, see?

  282. Kudlak says

    @ Narf #297
    You can see it, but you’re not like God’s girlfriend, unwilling to admit that she’s being abused least he dump her. To Christians, it’s all flowers and candy. Sure they know that they have to do exactly what he says, or else, and that they’ll never be anything like equal partners in the relationship, but they’ve been taught to believe that their whole life is pointless without a God/man to do for, so they’re willing to tolerate anything rather than be alone, like us.

  283. Kudlak says

    @Narf
    Frank was describing the instinctual drive to have leaders within our own species while I was describing how that drive led to the creation of gods who have to be of a different species than we are, with more power to see wrongdoing and punish. Enter the creation of the supernatural, right?

    TTFN

  284. frankgturner says

    @ Kudlak
    I would be hesitant to use the word “supernatural.” I would also be hesitant to insist that the god be of a different species so much as it have greater powers in order to protect those who worship it.
    .
    You have part of what I was getting at (the looking for leaders within our species part). I meant to imply that one can extend this to the idea that in the absence of leaders within our species who are powerful enough to protect us we might fill the hole. An imaginary being who is powerful enough to do so fills this need/hole.
    .
    To connect that with your idea, in order to be powerful enough to do so the being may have to have powers beyond that of a human being and because it is imaginary that is not such a hard leap to take.
    .
    Mind you this is all spoken hypothetically. The whole idea is to propose why many individuals need to believe in a god.
    .
    Also keep in mind (this goes with Xtianity), many seemed to have wanted this being to have human attributes as well as powers to protect. The ability to heal at will, rice from the dead, etc were desirable traits. So a lot of Xtianity seems to have been marketing strategies.

  285. Narf says

    @301 – Kudlak
    So, just a slight elaboration on what Frank was saying, then? It initially seemed like you were trying to reverse the metaphor.

  286. Narf says

    @Frank

    An imaginary being who is powerful enough to do so fills this need/hole.

    … and who might be convinced to protect us, if we provide tribute, as we did to pack alphas.

    The ability to heal at will, rice from the dead …

    Is that sort of like the reverse of spontaneous generation, with mice spontaneously generating from grain? God can do it the other way around?

  287. Monocle Smile says

    @Frank
    I get what you’re saying, but EL is most definitely not adamah. I’ve had correspondence off the blog with EL. He’s enterprising and friendly, and, you know, actually human, unlike adamah.

  288. EnlightenmentLiberal says

    @MS
    Aww, shucks. Thanks. Calling me human. That’s so nice.
    ~quickly dons his fake human skin.~
    NO ONE MUST KNOW!

  289. frankgturner says

    @ Matt
    That is supposed to be “rise” from the dead. Pardon the typo. And as far as willing to provide protection of we provide tribute, you are definitely getting where I was going.
    .
    Kudlak can elaborate on it. I have no objection to that. That is actually a good idea.
    .
    @ MS
    He definitely has been behaving like Adamah and I would hope that this sinks in that I am not the only one who sees that. The fact that the issue has been going on for some time with him and Matt makes me wonder if he ever really did consider that maybe people really don’t disagree but that it is just his perception of the situation. I would never argue that on some level any and all occurrences do influence probability of a situation. However, since we largely make estimations in life we tend to restrict what we state outright about said determinants to those occurrences which make significant enough influence upon the outcome to potentially change the outcome of the focus.
    .
    On some level we may be dealing with semantic differences regarding the words “consider”, “implicit” and “explicit”. After reading a bit of EL’s comments in this thread though it seemed like he is having imaginary conversations and arguments in his head that he is responding to along with the common line.
    .
    This did spark my interest in another conversation though, one that 8 have had with apologists and more conservative believers (some of this comes from my time as a liberal minded Catholic).
    .
    The Book of Job comes with a common Hebrew disclaimer at the beginning, “There once was a man.” Other scriptural books don’t have such disclaimers. Hence many apologists will argue that they were not written in such a fashion as to indicate that they were parable. Yet not all of life and what is said comes with a disclaimer. Many hypothetical situations don’t lead with a phrase like “speaking hypothetically.”
    .
    I would appreciate some thoughts on this.

  290. Kudlak says

    @frankgturner #302
    I’ll agree that it’s a fine line when it comes to gods. Humans have typically imagined just super-powered versions of themselves, at least in Western culture. The Greek gods had the same needs to eat, drink and have sex that ordinary people did. They ate ambrosia to maintain their immortality, just as it seems Adam and Eve could have with eating from the Tree of Life in the Garden. They walked and talked, also like God in the Garden. The line between god and human was much closer back then, and it appears that people could reach that level by taking the same thing that gave the gods their power.

    So, it seems that YHWH initially began as just another typical Mediterranean anthropomorphic god. It took the Babylonian exile to force Jewish imaginations into making him first global, and then omni-present. While he transcended these human aspects he still retained our emotions, expressing anger, jealousy and (supposedly) love recorded in scripture in ways we can easily identify as still human. Even today, while Christians love to argue that God is beyond our human understanding, they continue to express their understanding of his motives and values in human terms. To do otherwise would be to define him as something so alien as to be unpredictable and untrustworthy, and even the most illogical amongst them would find it difficult to justify calling something like “God”.

    The Greek gods also had problems similar to ours, and this shows one of the other reasons to have gods as people could look upon the myths for insight in how to solve their own problems. Legendary figures may serve the same purpose, and it might be that some legendary figures became so inflated that they became the anthropomorphic gods. We certainly continue to see how mere humans can be elevated to godhood. Maybe this is where there’s a missing link between your and my ideas?

  291. Kudlak says

    @frankgturner #307
    Parables are an interesting topic. I personally know a few Christians who insist that the parables that Jesus told must have been true stories, as he could not lie. This seems silly, especially considering how modern-day Christians love to dream up new parables to help illustrate their beliefs. How many people actually believe that the “Footprints in the Sand” story actually took place, after all?

    To me, the parables as basically in the same spirit as the classical myths. They may never have actually happened, but they were made up to teach some bit of wisdom about being human and relating to others. They may have been the forebears of modern fiction. After all, doesn’t a fictional character have to be relatable in order to be good?

  292. EnlightenmentLiberal says

    @Frank

    He definitely has been behaving like Adamah and I would hope that this sinks in that I am not the only one who sees that. The fact that the issue has been going on for some time with him and Matt makes me wonder if he ever really did consider that maybe people really don’t disagree but that it is just his perception of the situation.

    Or, it’s also possible that 5 random people gathered online might all happen to be wrong about something.

    If you want to play this game, I can cite Richard Carrier himself and some personal conversations I’ve had about this topic (proper Bayesian reasoning) and how he’s in full agreement with me. I can also cite several other people offline with whom I’ve had this conversation, just random people to you, who also agree with me about Matt’s arguments and my counter-arguments.

    (We have a much weaker agreement on the supernatural. We both agree intrinsic methodological naturalism is bullshit. I think “supernatural” is a completely useless term that has no sensible definition, but Richard thinks that there is a sensible workable definition. I politely disagree with him on that minor point.)

    It doesn’t help Frnak that you have displayed a complete ignorance of the Bayesian method from step 1, and have thus far refused to try to learn from me about what it really is. Instead, you’re arguing against strawmen of the position.

    I suggest you review the bottom of my post 230.
    http://freethoughtblogs.com/axp/2015/05/17/open-thread-for-episode-918-matt-and-john/#comment-609116
    You might learn a thing or two, and then we might be able to have an intelligent conversation.

  293. frankgturner says

    @ EL
    I understand the Bayesian method better than you might think.
    .
    Yes random people can collect and be wrong about things, it happens all of the time. What I am getting at is that when individuals observe something and come to similar conclusions, even if the conclusion turns out to be incorrect, most individuals at least give what they are thinking a slight but of thought.
    .
    Frankly I don’t care to discuss things with you aa I think that you are rude and closed minded. Furthermore you seem to be having discussions in your head and appear so insecure about your position that you have to beat down anyone who disagrees with you. (Which is implied as to why to need to tell people that they are wrong, how they are are wrong, and why they are wrong). It does not seem to be good enough to you to be correct about ideas in general. It is as though you need to be right about every little detail. I know people like that whose self concept is that if you are not right about everything then you are wrong about everything.
    .
    I can sympathise as I was once like that. I grew past it. If you could you might realise that people actually agree with your position more than you realize. I think that your low self esteem prevents you from seeing that.

  294. EnlightenmentLiberal says

    @Frank

    What I am getting at is that when individuals observe something and come to similar conclusions, even if the conclusion turns out to be incorrect, most individuals at least give what they are thinking a slight but of thought.

    Again, this is flatly ridiculous in context. Consider the amount of due caution you want me to show you. Do you want me to show the same due caution when I’m in a room with 5 young Earth creationists? Please.

    I understand the Bayesian method better than you might think.

    Based on your fundamental misunderstandings of its application as demonstrated in this thread, I do not believe you. In fact, I believe you are wrong.

  295. frankgturner says

    @ EL
    At least if I give people a certain impression about myself I consider what I did to give them that impression. They may be wrong in what led them to that conclusion but I do think about how they got there.
    .
    And fine believe what you want. You believe that Matt still does not understand you with regards to Bayesian reasoning and you make some good arguments as to why he might. I wonder why you have never flat out called the show and asked him. Matt seems to be friends with Richard Carrier, did you ask Richard Carrier if Matt comprehends that?
    .
    Do you really want to explore those beliefs with an opened and skeptical mind willing to accept evidence that might demonstrate that your hypothesis is incorrect? Do you really want to explore all competing hypotheses? It does not seem like it from my point of view. It sounds like you just want people to agree with you.
    .
    For one who is so adamant about alternate hypotheses, you seem awfully unwilling to consider alternate hypotheses regarding what people are saying.

  296. EnlightenmentLiberal says

    I wonder why you have never flat out called the show and asked him.

    I have on called into the show. Not on this exact topic of Bayesian reasoning, but the highly related topic of intrinsic methodological naturalism. Matt ended the call after 20 minutes or so because he didn’t want to talk about such esoteric points. I’ve been meaning to try again, but I’ve been lazy.

    Episode 896.
    http://freethoughtblogs.com/axp/2014/12/14/open-thread-for-aetv-896-matt-and-jen/
    It was a nice call, but not all that productive. Unfortunately, I did not do well, and/or Matt was too prejudiced against my position. I’ve been considering alternative ways of approaching the same problem.

    I might try again soon. I think I have two approaches that might make progress. 1- Argue that Thor the thunderbolter is an eminently testable claim, and if that’s testable, then intrinsic methodological naturalism is a do-nothing assertion. At best, it means the same thing as “science does not work on the unobservable – supernatural or natural”. 2- Argue that Matt believes that the desk in front of him is a natural object, and note that Matt has explicitly stated that it’s a procedural impossibility to use science to show that something is supernatural, and Matt has implied frequently that natural and supernatural is a true dichotomy, which means that Matt’s belief “the desk is a natural object” is unfalsifiable, and unfalsifiable scientific beliefs like that are bullshit.

    Matt seems to be friends with Richard Carrier, did you ask Richard Carrier if Matt comprehends that?

    In several of our conversations, I have related the argument to the best of my ability, and Richard agrees with my positions in the disagreement.

    Richard definitely agrees with me w.r.t intrinsic methodological naturalism. I can provide links in 5 seconds.

    Further, in several other places online, unprompted by me, Richard has correctly stated that an important takeaway of Bayesian reasoning is that you are always comparing claims to alternative claims and that it’s impossible to evaluate a claim in a vacuum; it’s impossible to evaluate a claim without concurrently evaluating the competing claims. Finding a citation on that will be a little more difficult, but I’ll do if it needed.

  297. EnlightenmentLiberal says

    Of course, for a citation for the last point, I’ve already pointed people towards Richard’s peer reviewed book on the topic, Proving History. I can whip out quotes from that later when I get home if desired.

  298. Monocle Smile says

    @EL

    Further, in several other places online, unprompted by me, Richard has correctly stated that an important takeaway of Bayesian reasoning is that you are always comparing claims to alternative claims and that it’s impossible to evaluate a claim in a vacuum; it’s impossible to evaluate a claim without concurrently evaluating the competing claims

    *sigh* we’re not actually arguing otherwise. You’re just totally missing what’s meant by “competing claims.” In the previous situation, the competing claim we’re examining is “Alice did not stab Bob,” not “Charlie stabbed Bob.” And again, using that claim as a reference says nothing about whether or not we actually believe it.

    I might try again soon. I think I have two approaches that might make progress

    I hope they both involve watching the past 8 or so episodes, because I’d hate for you to make an ass of yourself after Matt altered his position. I still don’t understand why you continue to harp on Matt for things he’s said in the past when you’ve made zero effort to check if he’s drifted from that. At the very least, he’s added the word “currently” to the thing about being unable to test the supernatural. He’s also said that upon being able to test the “supernatural,” we’d roll it into the natural, and thus “supernatural” may be an incoherent concept.

  299. frankgturner says

    @ MS
    Frankly I think I know why Matt stopped bothering. EL can’t seem to comprehend that the quote from Carrier and what Matt has said are not in disagreement*.
    .
    * I might say “not necessarily” as it is a matter of perspective, to some degree. I am surprised that it has never occurred to EL that the statements are not in disagreement and to explore why that might be the case.
    .
    I think EL got it into his head that Matt’s statement means something that it doesn’t mean. I am not sure if it would make any sense for him to watch the episodes. You can’t be a true skeptic if you refuse to acknowledge or can’t recognize competing evidence.

  300. EnlightenmentLiberal says

    @MS and Frank
    Earlier, John said this:

    You can’t ignore that though, even to make a point — it’s central to the entire argument. Evidence that implicates Charlie in and of itself does nothing to exonerate Alice.

    That statement is wrong. Can we agree to that?

    I assert that Matt frequently makes equivalent wrong-headed assertions. Can we agree to that? Do you need more citations that Matt has made this point? All we need to do is look up in this thread to see Matt saying otherwise.
    Ex: Quoting Matt:

    I’m sorry, but you’re just wrong in saying this…in part, because we’re NOT examining a question, we’re examining CLAIMS. The claim “some god exists” can be examined even if the claim “no god exists” cannot.

    It’s the same fundamental mistake. If you cannot examine the claim “god does not exist”, then you cannot examine the claim “god exists”. The meaning and truth of a proposition “a god exists” can only be described by contrasting it with another claim like “no (observable) gods exist”. If you cannot do that, then you have no claim either way. You’re just speaking jibberish. You’re not right, and you’re not even wrong.

    http://rationalwiki.org/wiki/Not_even_wrong

    This is what it means to think like a Bayesian. That’s what it means to think like a positivist. That’s what it means to think properly.

    Quoting MS:

    I hope they both involve watching the past 8 or so episodes, because I’d hate for you to make an ass of yourself after Matt altered his position.

    Maybe he’s changed his position w.r.t. intrinsic methodological naturalism, but we’re not talking about just that now. I’m also talking about his epistemological fail made in comments in this very thread, quoted above in this very post.

    Given that Matt made that statement, I am also highly dubious that he’s moved sufficiently far in the right direction on intrinsic methodological naturalism.

  301. Monocle Smile says

    @EL
    I wish you’d stop hoisting yourself up as the pinnacle of “thinking properly” and holding correct positions. It’s seeming like you’re doing that more and more.

    I did answer your Shawshank questions in 282, but you just ignored that post of mine.

    It’s the same fundamental mistake. If you cannot examine the claim “god does not exist”, then you cannot examine the claim “god exists”

    Firstly, that’s different from your quote of Matt in a rather important way, and you’re totally missing something here…the definition of god. Matt’s saying that if we define a god, then we can examine whether or not it exists. Meanwhile, the claim “no god exists” can’t be examined for every possible screwy, crappy definition of god.

    Given that Matt made that statement, I am also highly dubious that he’s moved sufficiently far in the right direction on intrinsic methodological naturalism

    /wanking motion
    Whatever. I’ve pretty much repeated myself for five or more straight posts, so I’m pretty much done with this thread.

  302. EnlightenmentLiberal says

    @MS
    Matt didn’t say “Well, we can’t examine either claim until we get a proper definition”. Matt clearly wrote in the same sentence that one claim can be examined and the primary competing claim cannot be examined. It is hard to fathom how Matt might have been clearer in this wrong position.

    I also do not understand how you can adopt such a forgiving interpretation of what John has said. That’s clearly not what John meant.

  303. EnlightenmentLiberal says

    @MS
    Further, the context makes it even readily more apparent that my understanding is correct.

    EL: You are wrong when you say that we can examine the question “is there a god?” without also examining the question “is it true that there are no gods?”.

    Matt: I’m sorry, but you’re just wrong in saying this…in part, because we’re NOT examining a question, we’re examining CLAIMS. The claim “some god exists” can be examined even if the claim “no god exists” cannot.

    He wrote that as a direct response and counter to my statement involving quote “is there a god?” and “is it true that there are no gods”. Matt clearly and unequivocally denied my position.

    You’re repeating yourself 5 times MS because you’re wrong and you’re not reading carefully enough. I admit that I was on auto-pilot for that one post, and I apologize, but for the most part you, Frank, and corwyn have been oblivious to the obvious and reaching for other ways you might interpret Matt and John to be anything but flagrantly wrong, and I have no idea why.

  304. Monocle Smile says

    @EL
    I didn’t make a statement about what John said, so now you’re just seeing things. This is getting a little ridiculous.

    You’re repeating yourself 5 times MS because you’re wrong and you’re not reading carefully enough

    You’re coming close to breaking my irony meter. You’ve haven’t actually addressed what I’ve said. You’ve mostly ignored it and focused on what other people said. Furthermore, I’ve never really, truly cared if Matt and/or John are actually right in what they’ve said before. You’re assuming that I do. When I mention them, I’m mostly playing Devil’s Advocate because I think you’re being grossly uncharitable. You’re not dealing with five young-earthers.

    He wrote that as a direct response and counter to my statement involving quote “is there a god?” and “is it true that there are no gods”. Matt clearly and unequivocally denied my position

    Actually, he was talking about questions vs. claims. Also, I deny your position as well, because you’ve set up a false dichotomy.

    but for the most part you, Frank, and corwyn have been oblivious to the obvious and reaching for other ways you might interpret Matt and John to be anything but flagrantly wrong, and I have no idea why

    So you’re not even going to consider the fact that you might be either totally misreading stuff and/or are the one who’s wrong. Fantastic.

  305. frankgturner says

    @ MS
    I keep wanting to bring up the film “My Cousin Vinny” and see if he gets the point. I seriously doubt it though.

  306. EnlightenmentLiberal says

    @MS
    I did consider that I misread it. That’s why I went back again and reread it, to make sure I’m in the right, and I am.

    Actually, he was talking about questions vs. claims. Also, I deny your position as well, because you’ve set up a false dichotomy.

    First, I don’t understand why you would jump through so many hoops to deny that’s what Matt meant, but only to say “but you’re still wrong even if Matt said that”. It feels like “moving the goalposts”.

    This is what I said:

    You are wrong when you say that we can examine the question “is there a god?” without also examining the question “is it true that there are no gods?”.

    Is this what you meant? Do you really mean that this is a false dichotomy? Really? Really? If so, I don’t know what to say. I missed something important. Please explain it to me.

  307. frankgturner says

    @ Kudlak #310
    Yeah Xtians like that who insist the parables must be factual because Jesus could not “lie” have probably never thought too deeply about what it means to “lie.” It is not like Jesus would put a disclaimer on a parable. Can you imagine that?
    .
    All individuals in this story are purely fictional. Any similarity to actual persons…
    .
    There is an episode of Red Dwarf where there is a news story where a hidden tablet from Genesis is found and the translation is basically a disclaimer.

  308. Monocle Smile says

    Is this what you meant? Do you really mean that this is a false dichotomy? Really? Really? If so, I don’t know what to say. I missed something important. Please explain it to me

    It’s either a false dichotomy or very poorly worded. “It is true that a god exists” and “it is not true that a god exists” is a true dichotomy. “True” and “not true” compose a true dichotomy. “True” and “false” do not compose a true dichotomy. This is 101 stuff.

    First, I don’t understand why you would jump through so many hoops to deny that’s what Matt meant, but only to say “but you’re still wrong even if Matt said that”. It feels like “moving the goalposts”

    I had hoped that this discussion had risen above “EL is right and Matt is wrong,” but I guess I was mistaken.

  309. EnlightenmentLiberal says

    @MS

    I had hoped that this discussion had risen above “EL is right and Matt is wrong,” but I guess I was mistaken.

    I would much prefer to focus on the contentions on the issues rather than on “he said, she said”. Let’s see if we can try that.

    “True” and “not true” compose a true dichotomy. “True” and “false” do not compose a true dichotomy. This is 101 stuff.

    Apparently I flunked 101. I think your pithy phrasing there is the best summary of everything wrong with the proper reasoning of most other people in this thread.

    You see, where I come from, in my logic 101 class, I learned that I can substitute “true” and “not false” for each other interchangeably, and that they mean the same thing. Same for “not true” and “false”.

    Further, in another 101 course, English, I learned that the following four sentences have distinct meanings:
    * I believe it is true.
    * I believe it is false.
    * I do not believe it is true.
    * I do not believe it is false.

    Further, in my science 101 course, I learned that it is either true that 1- there is a toy elephant on my table, or 2- it is true that there is no toy elephant on my table. Note that any impreciseness of the concepts “toy elephant”, “on” “my”, “table”, etc that apply to my formulation would equally apply to your formulation.

    I really don’t know what to say if you don’t think the toy elephant example above is a true dichotomy. If you do not see this readily obvious fact, then I do not think that rational discourse is possible.

    When I deal with young Earth creationists, sometimes they seriously propose that the devil put the fossils in the ground to fool them with basically no evidence to back that up. At that point in the conversation, I generally give up the conversation and declare intellectual victory for myself.

    I’m very close to doing that now with you. If I have successfully reduced you to saying sheer absurdities like “there’s a third option to 1- a toy elephant is on the desk, and 2- there is no toy elephant on the desk”, then I think I’ve “won”, or at the very least I see no reason to continue the conversation. I can only compare you to a creationist whom I’ve got stuck in a corner with no way out, forced to defend absurdities like “slavery is moral if god commands it”.

    I really hope that I’m misunderstanding you somehow, but I don’t have much hope on that matter.

  310. EnlightenmentLiberal says

    PS:
    @MS
    I think many others, you’re confusing the belief space and the event space. IIRC, even Matt would agree that there either is a god, or there isn’t, and from where I come from, that has the same meaning as it is true that there is a god (or gods), or it is true that there are no gods. Again, Matt would be quick to point out that our beliefs are not the same thing as what’s actually true. Matt like’s the comparison of maps. IIRC, it’s even a formal logical fallacy of confusing the map of the place for the place. Please don’t confuse your beliefs about reality for reality.

    http://en.wikipedia.org/wiki/Map%E2%80%93territory_relation

    In reality, it’s either true that there is a toy elephant on my desk, or it’s true that there is no toy elephant on my desk. However, our beliefs regarding reality on this point is not binary. We might be convinced that there is a toy elephant. We might be convinced that there is no toy elephant. We might be undecided.

  311. Monocle Smile says

    @EL
    The rest of us are talking about belief space. Are you talking about event space? That explains a few things. This discussion has always been about belief space, at least I thought it was. When did we switch to event space? Because for the most part, I don’t really care to discuss event space; it often leads to red herrings about absolute certainty.

  312. EnlightenmentLiberal says

    @331

    The rest of us are talking about belief space. Are you talking about event space? That explains a few things.

    Lol. I’ve only said it like a bazillion times in the thread.

  313. Monocle Smile says

    What? Where? I don’t recall that being explicit. This discussion started out about beliefs, and several of your posts even go towards “drawing conclusions,” which concerns belief space. I guess I just wasted a lot of effort, given my apathy towards discussing event space in these kinds of discussions.

  314. EnlightenmentLiberal says

    MS:

    The rest of us are talking about belief space. Are you talking about event space? That explains a few things.

    EL:

    Lol. I’ve only said it like a bazillion times in the thread.

    MS:

    What? Where? I don’t recall that being explicit.

    Oh, I don’t know. Maybe just in all of the following posts of mine?

    I think I was more than sufficiently clear in several earlier posts, but I will include only the posts where I’m “beating a dead horse” clear.

    #190

    I have no idea what you’re talking about. Either the person did the crime, or they didn’t. There’s no in between. Our estimations may be in between 0% and 100%, but the facts of the matter are that they did it or they didn’t.

    #191

    If you are a betting person, you need to calculate the worst payout bet that you would be willing to accept when someone offers you a bet, and that depends on your estimation of the likelihood that you will win the bet. Again, suppose someone claims incontrovertible proof of innocence and guilt, but offers you a bet of $X net payout for innocence, and $Y net pay on guilty. In order to determine if you want to take that bet, you need to estimate the probability that you will that bet. You will arrive at a certain minimum betting odds that will be willing to take, for example 1 to 2 odds.

    Suppose the person offers you a another bet, a reverse bet, where they offer to pay you some money on guilty, and you win some money on innocence. Your earlier calculation will necessarily constraint what bets you would be willing to take.

    To put it one way, Bayesian reasoning is all about giving betting odds for the truth of propositions. And if I’ve done my fourth grade math right, that means if you initially were willing to go no worse than X to Y odds, it means that you should be willing to go no worse than Y to X odds for the reverse bet.

    #198

    Also, There is a possible confusion on your part in terms. Let me repeat myself again. There is the state of affairs “not guilty”, and there is the verdict “not guilty”.

    As a working assumption for the sake of argument, either someone committed the crime or they didn’t. They are either innocent or guilty. They are either guilty or not guilty. They are either innocent or not innocent. This is simply a description of the event space. The set of all events which are not the guilty event equals the set of the innocent event.
    event space = { “guilty event”, “innocent event” }
    { X in event space : X != “guilty event” } = { “innocent event” }

    The “not guilty” verdict is not a description of the event space. The “not guilty” verdict is a description of the jury’s confidence about the truth of propositions “X is guilty”, “X is innocent”, “X is not guilty” and “x is not innocent”. In particular, a “not guilty” verdict implies the following Bayesian confidence levels (assuming 99% cutoff for “beyond reasonable doubt”):
    P(guilty) in the range [0%, 99%]
    P(not guilty, the event, not a the verdict) in the range [1%, 100%]
    P(innocent) in the range [1%, 100%]
    P(not innocent, the event) in the range [0%, 99]

    In context, P(not guilty the verdict) makes no sense. The thing that goes in the P() is an event in the event space. “Not guilty the verdict” is a description of confidence levels, not an event in the event space. You’re confusing descriptions of confidence with events in the event space.

    #205

    I don’t know what else to say. I have explained your errors – in short, you’re confusing “not guilty” as a description of the event space and “not guilty the verdict” which is a description of confidence levels. You’re seemingly hopelessly confused, and wrong.

    #206

    […] do you agree that the event space in the real world for a criminal trial can be accurately modeled as { “guilty”, “innocent” } ? Am I missing anything? I don’t know what it means to disagree. It’s incoherent.

    Can we talk about the sub-space of events described with the English “not guilty” ? What would that look like? In set-builder notation, I say it looks like:
    The set of “not guilty” = { X in event space : X is not “guilty” } = { “innocent” }
    Can you agree to this? If someone is actually not guilty of a crime, that means they’re innocent of the crime, right?

    @207

    Again, please see the difference between “event space” and “my estimation of the truth of the matter, and my beliefs about the truth of the matter”. The event space is { “innocent”, “guilty” }. Your belief about the event can be anything in the range: “highly confident innocent”, “leaning innocent”, “undecided / agnostic”, “leaning guilty”, “highly confident guilty”, and everything in between. A “not guilty verdict” includes the estimations { “highly confident innocent”, “leaning innocent”, “undecided / agnostic”, “leaning guilty” } and excludes “highly confident guilty”. Again, you are confusing the event space – how the world actually is – with your knowledge of the world. The map is not the place.

    #230

    You’re confusing events in the event space with beliefs in the belief space. […]

    A full length dissertation follows that explaining the difference. (/joke)

    #264

    I say that the proper way to think about problems like this is to break the space of epistemically plausible / possible real world events, the event space, down into mutually exclusive events that cover the entire space (or at least down into events that are well defined with well defined overlaps). The idea that you can consider the claim “Alice stabbed Bob” without positively considering the claim “Charlie stabbed Bob” is wrong and outright foolish. Matt, John, et al would disagree, and I’m still at a loss how. I’ve tried to explain, but their replies indicate a near complete lack of understanding of everything at hand.

  315. Monocle Smile says

    Ah. So it happened mostly in the vast exchange between you and frank that I ignored because I couldn’t find a way to care. Guess I deserved that one. I still stand behind my comments before my blunder.

  316. frankgturner says

    I was hoping not to have to do this, largely in part from the idea that individuals get more out of something that they took it upon themselves to do rather than by being prompted to do so. Young earth creationists are not motivated to actually study biology or geology (with a few exceptions) in an effort to determine why evolution has such a strong backing. Some apologists claim to have significant knowledge of, for example, evolution, but obviously don’t (WLC comes to mind).
    .
    As such I largely ignored the “event space” comments, particularly with regard to {innocent, guilty} of the crime in question. It would be more accurately described as {innocent of action, guilty of action} as the word “crime” makes certain assumptions. That is where The legal term “actus reus” comes into play. The toy elephant analogy is good as it is describing a true dichotomy.
    .
    I was hoping that without promoting EL he might look up some legal terminology. I considered just telling him in the same condescending manner in which he tells others but then he might be resilient to actually looking up legal terms.
    .
    What I largely don’t think is productive is his attitude. The concept of “declaring intellectual victory” about young earth creationists indicated a type of “superior” attitude (if we were religious I would call it “holier than thou,” to use a metaphor). This whole “appropriate” or “correct” way of thinking is what evangelists and apologists do and it is a lot of bullshit. Unfortunately EL here may have dropped the religious beliefs but he did not drop the superior condescending attitude. I think that might be why the hosts don’t bother engaging with him, and many others. He is lucky that some of us have had the patience to put up with this.
    .
    So to MS, whereas he may be right in his thinking in terms of logic and factual correctness, this never did move past the “EL is right and Matt is wrong” sentiment because EL here is too lacking in self esteem to get beyond that. Frankly I think he is too lacking in self esteem to actually look up legal terminology because it might demonstrate that he is full of shit. Much like young earth creationists are too lacking in self esteem to make an effort to look up evolutionary biology and understand it. They are too afraid of being wrong and I think EL is afraid of being wrong.
    .
    So the fact that you did not seem to much care about event space vs believe space MS is understandable. I knew and it made sense, I just did not acknowledge it given the attitude of the individual talking about it.
    .
    Frankly I think EL’s attitude about debate and there being a “winner” and a “loser” and someone who is “right” and “wrong” needs tempering. Debates can be about a sense of understanding each other’s view and trying to learn from each other. That is why you try to make your opponent’s argument for them, it gives you a sense of understanding.

  317. corwyn says

    @336 Frank:

    A person interested only in the ‘event space’ doesn’t talk about ‘guilty and innocent’ as those are description only useful in a *legal* space, not event space.

  318. frankgturner says

    @ corwyn
    Exactly, which is why I thought the wording was poor and that hopefully an individual would recognize that people who object to it have valid reason to do so IF one makes an effort to educate oneself on the legal terminology. I used the example “innocent” of action vs “guilty” of action for lack of better terms. (If someone has a better way of expressing this I would acknowledge it).
    .
    I would have just outright stated something to that effect but I did not want to prompt as I thought it would be better to let someone look that up on their own. I guess that I was wrong and I should have just flat out stated it.
    .
    I almost posted a several equations like the one you did corwyn several times, particularly with regard to how when we are calculating the probability of the event and how we are going to make a verdict is an “estimate” of how the belief maps to the event taking place and how evidence is often lacking.
    .
    I didn’t just want to give a lecture on law though as I thought it more advantageous for a person to look it up on their own without prompting. I was wrong.

  319. Kudlak says

    @ frankgturner #327
    Acting use to be widely considered sinful in the 1800s because it basically entailed “lying” to an audience. This association was made long before there was a Hollywood.

  320. John Iacoletti says

    @MS and Frank
    Earlier, John said this:

    You can’t ignore that though, even to make a point — it’s central to the entire argument. Evidence that implicates Charlie in and of itself does nothing to exonerate Alice.

    That statement is wrong. Can we agree to that?

    I stand by it. You can’t just assume that one and only one person killed Bob. Evidence for Charlie killing Bob is examined at Charlie’s trial, and evidence for Alice killing Bob will be examined at Alice’s trial.

  321. frankgturner says

    @ Kodiak #339
    Yeah I read that somewhere and I consider it pretty foolish that people would never think about intent to deceive, but some people are just dumb. I once spoke to a Jehovah’s Witness who was totally dumbfounded by the idea that a story could carry wisdom and meaning without it being “true” (“factually correct”). He felt even more foolish when I have him an example of him having done exactly that previously.
    .
    @ John #340
    I think we are dealing with some discrepancy over “formal consideration” vs what one might call “ideal” consideration. In an ideal world we have all of the details so every little bit of evidence factors into our equation. That is basically the “event space.” ANY hypothesis, no matter how vague, is factored in as a competing hypothesis. Many competing hypotheses have so little evidence supporting them that they influence the math insignificantly, but to say that they “don’t influence the math “at all” would be dishonest. In that regard all competing hypotheses DO influence the calculation. So whether or not Charlie did the act does influence the calculation of whether Alice did it. As does the probability of EVERY other person in the world doing it instead of Alice.
    .
    In the real world we don’t have that. We won’t have enough evidence to examine EVERY competing hypothesis. So what is being examined is a claim, which in this case is a belief as compared to the event itself. A formal charge is essentially a statement of belief. The formal charge is usually accompanied by a hearing to determine if the evidence collected is sufficient enough to conduct a trial, which is (among other things) a review of the evidence to determine if the evidence presented is sufficient enough to convince a group of individuals if this evidence maps this belief to reality. It can be wrong, sometimes evidence which supports a belief does not map to reality.
    .
    Of course there are other things determined by trial, such as if the action should be considered a crime. What effectively happens is that the terms “crime,” “innocence,” and “guilt” are all highly conditional terms. A LOT of assumptions are made in legal terminology regarding these terms.
    .
    When a formal statement of belief is made regarding Alice’s guilt in the situation, all the defense is driven to do is show that the evidence does not support this belief. It could be because Alice did not do the action, what occurred was not a crime, Alice is not in her right mind, etc. (some of these overlap).
    .
    So to determine if that belief is supported by evidence, no Charlie’s guilt need not be established to come to the conclusion that Alice should be exonerated of guilt. It helps, having an alternate explanation supported by evidence can help in determining if a belief is supported or not supported (by showing support for the compliment) by evidence.
    .
    The ACTION on a certain date and time is restricted to a more dichotomous space. It either did or did not occur. In that regard Charlie’s guilt DOES influence the calculation of whether Alice did the action. As do an infinite number of other explanations.
    .
    The court is examining not only if the action occurred, but whether it should be considered a crime. That is based on belief and conditional statements. The amount of evidence (basically everything) taken to determine that is near impossible for the court to consider, particularly when only a small subset of beliefs (Alice’s guilt or non guilt) is being formally considered.
    .
    So when considering a belief about Alice’s guilt or non guilt, small sub sets of evidence that can be obtained are analyzed. Basically what is being calculated with regard to that belief claim is an estimate of the actual event space (that probably WON’T consider all competing hypotheses unless they can influence the verdict).
    .
    I may have worded some of that poorly but it gets into what a court determines and I have an extension of this as to why that makes sense to use as an analogy for beliefs in the existence of God.

  322. EnlightenmentLiberal says

    @John Iacoletti
    I ask again: In The Shawkshank Redemption, do you think that Andy’s fellow prisoners were being irrational when they changed their minds about Andy’s guilt and innocence upon hearing the hearsay confession?

    @Frank

    The court is examining not only if the action occurred, but whether it should be considered a crime. That is based on belief and conditional statements.

    Sometimes yes. Sometimes no. Not all crimes require intent, “actus reus”. There are plenty of strict liability laws, and there are plenty of common law precedent for criminal negligence. It’s flatly silly to blanket assert that you always need actus reus for conviction. Oftentimes, merely proving physical actions is completely sufficient for criminal conviction.

    @Frank and Corwyn
    Again, I come from a different world where English is used differently. In the world that I come from, in the common parlance, the English statement “that person is guilty of some particular crime” is largely synonymous with the English statement “that person committed some act which has been outlawed, with any necessary conditions such as actus reus, and without any exonerating conditions such as justified self defense”.

    In the world I come from, there is a difference between “The person is not guilty” and “The jury has delivered the verdict ‘not guilty’.” The first is an expression of confidence of innocence, and the second is a technical legal verdict which is the result – the expression of the jury’s confidence of guilt not exceeding a “reasonable doubts” threshold.

    http://en.wikipedia.org/wiki/Strict_liability_%28criminal%29
    I don’t care that much if you’re condescending, but be right. If one is not right, then I don’t care if one is condescending or not – one would still be wrong.

    @corwyn and frankgturner

    A person interested only in the ‘event space’ doesn’t talk about ‘guilty and innocent’ as those are description only useful in a *legal* space, not event space.

    Exactly, which is why I thought the wording was poor and that hopefully an individual would recognize that people who object to it have valid reason to do so IF one makes an effort to educate oneself on the legal terminology.

    In the world where I come from, in the common parlance, “guilty” is interchangeable with “actually performed the actions which are outlawed, along with any necessary intent, planning, etc., along with no sufficiently excusing reasons like self defense, etc.”. I recognize that “guilty” is also a legal vedict delivered by a jury.

    In the world where I come from, there is the legal verdict “not guilty” which is distinct everyday “not guilty”. For example, in that world, “That person is not guilty” has a different meaning than “The jury found the person ‘not guilty’.”

    @Frank

    So to determine if that belief is supported by evidence, no Charlie’s guilt need not be established to come to the conclusion that Alice should be exonerated of guilt. It helps, having an alternate explanation supported by evidence can help in determining if a belief is supported or not supported (by showing support for the compliment) by evidence.

    I think I agree in large part, but I think I disagree in a significant part. It is not necessary to demonstrate Charlie’s guilt, or anyone else’s guilt, in order for the jury to properly arrive at the legal verdict “not guilty” for Alice. However, it is required that there be some explanation (or explanations) of the available evidence alternative to “Alice is guilty” in order for the jury to properly arrive at the legal verdict “not guilty” for Alice. You don’t need to demonstrate that this alternative is true beyond reasonable doubt. You don’t need to demonstrate that this alternative explanation is more likely true than false. You just need to demonstrate enough plausibility of this explanation so that it introduces reasonable doubt on the claim that Alice is guilty.

    So when considering a belief about Alice’s guilt or non guilt, small sub sets of evidence that can be obtained are analyzed. Basically what is being calculated with regard to that belief claim is an estimate of the actual event space (that probably WON’T consider all competing hypotheses unless they can influence the verdict).

    You lost me a bit here. I’m not sure what you’re saying.

    A proper jury needs to estimate the likelihood of truth of all of the events in the event space. Of course, even listing out all of the events in the event space may be an impossible task, and so that’s when our heuristics kick in, and we first narrow down the event space to plausible events, and then we estimate the likelihood of the truth of all of those events. The sum of those estimations should total 1, e.g. 100%, otherwise you did something wrong.

    Then, the jury should sum up all events that qualify as “Alice is actually guilty”, and if that number exceeds the reasonable doubts threshold, which maybe is 99%, then the jury should deliver the legal verdict “guilty”. Otherwise, the jury should deliver the legal verdict “not guilty”.

    PS: A person’s beliefs and intents are real facts about the real world, and they compromise events in the event space. “Alice killed Bob, but it was an accident” is a distinct event from “Alice killed Bob with premeditation”.

  323. Monocle Smile says

    In the world I come from, there is a difference between “The person is not guilty” and “The jury has delivered the verdict ‘not guilty’.” The first is an expression of confidence of innocence, and the second is a technical legal verdict which is the result – the expression of the jury’s confidence of guilt not exceeding a “reasonable doubts” threshold

    I find it a bit disingenuous to pretend that the first is of any relevance when the whole point of the analogy is purely to talk about the second. That’s literally what we’ve been talking about with regards to the analogy the entire time.

    It is not necessary to demonstrate Charlie’s guilt, or anyone else’s guilt, in order for the jury to properly arrive at the legal verdict “not guilty” for Alice. However, it is required that there be some explanation (or explanations) of the available evidence alternative to “Alice is guilty” in order for the jury to properly arrive at the legal verdict “not guilty” for Alice. You don’t need to demonstrate that this alternative is true beyond reasonable doubt. You don’t need to demonstrate that this alternative explanation is more likely true than false. You just need to demonstrate enough plausibility of this explanation so that it introduces reasonable doubt on the claim that Alice is guilty

    Thanks for finally agreeing with what I’ve been saying thus far, at least sort of…as long as those “explanations” include those pulled out of a hat and assuming that you’re abiding by the definition of “evidence” as “data that fits one hypothesis better than its competitors,” because plenty of facts about the case likely won’t fit any hypothesis better than any other, like Alice and Bob driving the same make of car.

  324. EnlightenmentLiberal says

    EL: In the world I come from, there is a difference between “The person is not guilty” and “The jury has delivered the verdict ‘not guilty’.” The first is an expression of confidence of innocence, and the second is a technical legal verdict which is the result – the expression of the jury’s confidence of guilt not exceeding a “reasonable doubts” threshold

    MS: I find it a bit disingenuous to pretend that the first is of any relevance when the whole point of the analogy is purely to talk about the second. That’s literally what we’ve been talking about with regards to the analogy the entire time.

    I don’t understand. I need to talk about the event space in order to explain proper Bayesian reasoning which is the method by which you form proper beliefs and confidence levels of those beliefs.

    I felt it prudent and reasonable to say “Either the person is guilty, or he’s not. Either the person did the crime, or he didn’t. Either the person is guilty, or they’re innocent.” I felt it prudent and reasonable to do so to emphasize that there’s the event space on the one hand, and there’s our beliefs about the event space on the other.

    I’ve re-read the entire thread several times now, and I still think I was more than sufficiently clear even from the beginning. Still, there has been a massive failure in communication in here somewhere, and it might still be me, partially or wholly.

    MS: Thanks for finally agreeing with what I’ve been saying thus far, at least sort of…as long as those “explanations” include those pulled out of a hat and assuming that you’re abiding by the definition of “evidence” as “data that fits one hypothesis better than its competitors,” because plenty of facts about the case likely won’t fit any hypothesis better than any other, like Alice and Bob driving the same make of car.

    Glad me made progress. Sorry that I’m seemingly a very poor communicator.

  325. John Iacoletti says

    @John Iacoletti
    I ask again: In The Shawkshank Redemption, do you think that Andy’s fellow prisoners were being irrational when they changed their minds about Andy’s guilt and innocence upon hearing the hearsay confession?

    I haven’t seen it. But if I can generalize, I would say that this is irrelevant because we’re talking about a jury finding someone not guilty in a legal sense.

    In the world I come from, there is a difference between “The person is not guilty” and “The jury has delivered the verdict ‘not guilty’.” The first is an expression of confidence of innocence, and the second is a technical legal verdict which is the result – the expression of the jury’s confidence of guilt not exceeding a “reasonable doubts” threshold.

    I think we’ve finally hit upon the crux of the disagreement. As MS pointed out, we’ve been talking about the second meaning throughout all of this. The whole point of the courtroom analogy is to compare “not believing in a god” to “not believing in a person’s guilt” which has a different standard of evidence than believing there is no god or believing a person is innocent. If you’re a priori defining “not guilty” as “an expression of confidence of innocence”, then this becomes a circular argument because that’s what you’ve been trying to prove with your Bayesian analysis. You can declare that your pre-assumption is “correct” and others without that pre-assumption are “wrong”, but that’s not particularly enlightening to either side.

  326. EnlightenmentLiberal says

    @John
    In the world I come from, the following sentences have different meanings.
    * I believe that X is false.
    * I do not believe that X is true.

    1 Can we agree to that?

    In the world that I come from, the following sentences have the same meaning.
    * I believe that X is not true.
    * I believe that X is false.

    2 Can we agree to that?

    Now, I know that it’s a common thing in the modern English language to confuse these two statements. I know it’s common shorthand to say “I do not believe X” when one really means “I believe that X is false”. However, I’m trying to be precise here, and I think I have been more than precise.

    In the world where I come from, for a simple well-formed boolean proposition, either it’s true or it’s false. There is no third option. Our knowledge of the truth of the boolean proposition may be incomplete. It’s not right to say that one’s belief must be “it is true” or “it is false”. However, we should all be able to agree that in actuality it is either true or false, but we might not have enough evidence to safely conclude it’s true nor to safely conclude it’s false.

    3 Can we agree to that?

    Similarly, in the world where I come from, in the common vernacular, for a particular person and a particular crime, either they committed the crime or they didn’t. Either they are guilty of the crime or they’re not. Either they’re innocent of the crime or they’re not. Either they’re innocent or guilty. Our knowledge of guilt and innocence is not binary. Similarly, our verdicts of “guilty” and “not guilty” represent our incomplete knowledge about the events and represent our desire to only convict people for which we have confidence beyond reasonable doubt.

    4 Can we agree to that?

    Ok. Let me describe the scenario to you. Andy is convicted of the murder of his wife and the wife’s lover. Andy is placed in prison. Years later, a new prisoner enters the prison named Tommy. Tommy shares a story of a third person who was his previous cellmate who often bragged of killing people. Tommy shares a story of this one time this unnamed cellmate bragged that he killed a married woman who was cheating on her husband with another guy, and killed the guy, and how the police blamed the innocent banker husband for the crime. With a few other minor details, they could confidently narrow down the story to Andy’s dead wife and lover. This does not necessarily convince the other prisoners that Andy is innocent, but it definitely changes their estimations of Andy’s guilt, and they decide to work together to get a re-trial for Andy on the basis of this new information.

    Do you agree with the reasoning of Andy’s fellow prisoners? Do you think that they did proper reasoning when they learned of evidence (really bad evidence, hearsay) that someone was committed the crime, and from that significantly lowered their estimation of the likelihood of Andy’s guilt? If you were a police investigator or prosecutor learning of this, wouldn’t you want to go find that unnamed cellmate and see if you can corroborate the hearsay? Wouldn’t this hearsay adjust your estimation of the likelihood of guilt?

    If you think the hearsay component is too big of an issue, what if Tommy himself confessed to the crimes, seemingly honestly, with information that he could only know if he committed the murders. As one of Andy and Tommy’s fellow prisoners, shouldn’t that make you lower your estimation of the likelihood of the guilt of Andy for the murders?

    It should be obvious that learning of this hearsay or direct confession should adjust your estimation of Andy’s guilt at least to some degree. I’m not saying it should convince you of Andy’s innocence. I’m not saying it necessarily should make you no longer vote for a guilty verdict. I’m not saying it necessarily should make you vote for a “not guilty verdict”. I’m just saying that this should make you lean away from the guilty verdict and more towards the not guilty verdict. Maybe it’s not enough on its own. Maybe it is. But it is evidence which you should take into account as a good juror, and it does force you closer towards a not guilty verdict.

  327. Monocle Smile says

    @EL
    I agree with everything in that last post…but I don’t think anyone would disagree. I’m glad that was all laid out, but none of those points are in contention. Once again, the point of dispute is something else. I refer back to my “necessary vs. useful” posts.

  328. frankgturner says

    @ EL
    For starters, in context you appear to have confused “actus reus” with “mens rea.” and in cases of criminal negligence the negligence is effectively being used as a substitute for “mens rea.” in some cases the action is enough to establish mens rea without it being formally examined. Doing some reading on mens rea though would let you know that it often IS examined as evidence that the actions committed were not considered criminal. An act of self defense defender defense of another that requires the taking of a life to protect an individual in imminent danger is Not considered a crime. However, you are not considered “innocent” of the action, you are considered “not guilty” of a crime (in common parlance perhaps but in legal terms no).
    .
    Also, given that you had several of us in here who comprehended why in legal terms “innocent” and “guilty” are not considered a dichotomy, why were you not driven to look up anything about legal terminology? That principle is near essential to understanding why the analogy is used for belief in God and you seemed so driven to “be right” rather than to reach an understanding.
    .
    Did you learn nothing from talking to young earth creationists?

  329. EnlightenmentLiberal says

    @MS
    I bet people still will disagree. Thankfully not you. That hunch might just be part of my miserable failure at communication.

    @Frank
    I cited two things: strict liability laws, and criminal negligence laws. You seem to have addressed only one of those things. With strict liability laws, the mental state of the person is largely irrelevant. That’s the defining quality of strict liability laws.

    Also, given that you had several of us in here who comprehended why in legal terms “innocent” and “guilty” are not considered a dichotomy, why were you not driven to look up anything about legal terminology? That principle is near essential to understanding why the analogy is used for belief in God and you seemed so driven to “be right” rather than to reach an understanding.

    I’m arguing against that because I think you are confused, both with legal guilt and with the existence of gods. I also see certain peculiarities with some people here, especially the hosts of the show, – peculiarities with this language when talking about the existence of god which suggest to me errors in reasoning. I would cite earlier quotes, but that was woefully unproductive, so I’m going to try to just move forward with you.

    I’m continuing to get the distinct impression that you Frank, and John, are still disagreeing with the core principles of Bayesian reasoning which I laid out in post 347 whereby finding evidence of guilt for one person often is evidence for non-guilt of another. I’ll have to wait for John to reply to see what’s going on there, but you didn’t engage the substance of 347 enough for my tastes, and I don’t know your position on it.

    I’ll try to work with you. You see, I have this concept in my head which I can describe as “a particular person has this property if and only if a proper jury, in possession of all of the relevant facts and with no uncertainties or ignorance as to the material facts, properly gives the guilty verdict on that person for a particular crime, and no jury nullification”. I have this related concept in my head which goes “a particular person has this property if and only if a proper jury, in possession of all of the relevant facts and with no uncertainties or ignorance as to the material facts, properly gives the not guilty verdict for a particular crime, and no jury nullification”. These concepts of mine relate to the world as it actually is, and not to our mere uncertain beliefs about the world. Specifically, the concepts operate under the assumption of perfect knowledge, and that’s why it’s a measure of reality and not of our beliefs about reality. Usually, I might describe these concepts with “he did it” e.g. “I assert positively that he did it” vs “he didn’t do it (or lacked a required mental state, or had a sufficient legally exculpatory excuse, etc.)” e.g. “I assert positively he didn’t do it”. Usually, I might describe these concepts with the words “guilty” and “innocent”. If it’s really the words that are getting to you, I can invent new words like “foo” and “bar”. Someone is foo if they committed the illegal act, and had any required mental state, and with no legally sufficient exculpatory excuse, etc. Someone is bar if they did not commit the illegal act, or if they lacked a required mental state, or if they had a legally sufficient exculpatory excuse, etc. Would that make you any happier? Do you think there’s a noteworthy difference between the words “foo” and “bar” as I have just defined them, and the everyday normal concepts of “guilt” and “innocent”? Or the technical legal meanings?

    I assert that the facts of the matter are that a particular person with respect to a particular crime is either foo or bar. There is no in between. Now, our knowledge about the real world and whether the person is foo or bar is always incomplete, and thus there is always uncertainty in our beliefs about the foo-ness and bar-ness of the person with respect to the particular crime. For example, we might be 90% convinced that the person is foo and 10% convinced that the person is bar. We might be completely undecided and ignorant which means a 50% confidence of foo and and a 50% confidence of bar.

    Is it any more acceptable if I use the words “foo” and “bar” instead of “innocent” and “guilty”? Note that I’ll probably have to include a legend if only for my own sanity, ex:
    Reminder: Loosely, I have defined “foo” to approximate “guilty” and “bar” to approximate “innocent”.

  330. frankgturner says

    @ EL
    I did not address the “strict liability” laws as I considered it largely irrelevant to the conversation at hand. I can address it in how sometimes “mental state” is about knowing what NOT to do which can be considered a “mens rea” if you read more about the purpose for why “mens rea” is considered, but it does not matter. You finally seem to be comprehending why the terms “innocent” and “guilty” are not considered strictly complimentary in legal terms. If you had been more willing to entertain THAT thought earlier on rather than insisting that they were things could have been a lot more productive. So yes, sometimes it is important to not just “be right” and to not give a shit if you are arrogant about it. You don’t have to invent alternate terms for perfect innocence and perfect guilt, just be attentive that the words are used differently in different situations.
    .
    If you really mean to try to identify the use of those words in such a way as they won’t be ambiguous, think of an adjective that you could add to them to be clearer (notice how I just defined what you described as “perfect” innocence and “perfect” guilt). Or alternately try to define the terms as you basically did. At least you are being attentive to their potential ambiguity.
    .

  331. frankgturner says

    @ EL continued
    Well now that you are demonstrating that what you have is an “impression” as compared to being certain that people are “wrong headed” I will let you know. Yes I recognize that the guilt of another person does contribute, rather significantly, as evidence of the non guilt of another. What I suspect people had the impression of here was that you insisted that you MUST demonstrate, in no uncertain terms, the “guilt” of another party to arrive at the conclusion that the accused party was “not guilty.” THAT would show a distinct lack of understanding about law.
    .
    Let me explain something with regards to

    in possession of all of the relevant facts and with no uncertainties or ignorance as to the material facts,

    When many students first go to law school one of the first things they do is demonstrate a potential crime by play acting a crime onstage. It varies from law school to law school. One about a man in tweed clothes with a hooded jacket wearing a face mask who comes into the room and stabbing another who is sitting playing a piano wearing a tuxedo is common. All of the students write reports down what they just saw and afterwards there are comparisons. Despite everyone having just witnessed the same thing (which is being filmed) they get TONS of differences in what their reports say including everything from the weapon used, what people were wearing, what crime was actually committed, what got said, etc. Many of those reports are compared to the video (people get a number of things wrong). This just gives them a TASTE of what law is about.
    .
    When a crime (or potential crime) is committed a LOT of things can be left behind and collected as both physical and mental evidence. Footprints from shoes, fingerprints, blood, bullet holes, etc. as well as emails and letters of plans and so on. The list of potential evidence is near infinite. When one has overwhelming quantities of evidence those cases tend not to go to trial. (Some do though).
    .
    In cases that tend to go to trial people will likely NOT have all of the relevant facts and there WILL be uncertainties and some ignorance of material facts that were not collected. So they have to make decisions based on what evidence they DO have. A lot of potential hypotheses are NOT considered because one does not have evidence to support them (even if they evidence exists it may not have been collected or its integrity may be in question). Sometimes evidence is specifically disregarded due to questions about its relevance or integrity. You might have recognized that in The Shawshank Redemption (mind you that is fiction but such things do happen) he gets convicted despite a large LACK of physical evidence due to his having mens rea.
    .
    On the same note, what is being formally considered is the guilt or non guilt of the accused. The potential guilt or non guilt of others is often presented in many cases as it does influence the consideration of guilt of the accused, but the other individuals may not be on trial (sometimes multiple people are formally charged). The relevance of how much the guilt of another will influence persuasion as to the guilt of another is part of the trial (if it is of low significance it is sometimes disregarded).
    .
    So a courtroom is NOT a place of pure Bayesian reasoning. It uses some ELEMENTS of Bayesian reasoning. A BIG part of the reason why it is not a place of pure Bayesian reasoning is the recognition that you won’t have all of the relevant evidence and there will be uncertainties. Even if you have overwhelming evidence you still do not have EVERYTHING. So what gets examined in these cases is an ESTIMATE of all potential hypotheses as suggested by the evidence presented.
    .
    Evidence may come along at a later time that influences the outcome, sometimes even after a verdict has been declared. The ability to test human DNA and use it as a type of “fingerprint” was not available for years and now that it is it is being used to re-examine a number of cases. Hopefully you have heard of the “Innocence Project” (they would call it the “Not Guilty Project” in proper legal vernacular but part of the idea was to appeal to the public who is unfamiliar with the difference in usage and to whom that could have been confusing).
    .
    In a court we recognize that we have incomplete information in most cases (one might say that unless we are omnipotent we have incomplete information in ALL cases as no case is perfectly ideal) and we are always opened to new information. THAT is a big part about why the analogy is used, i.e.: is god “guilty” of existing vs “not guilty” of existing based on our definition of god at the time and given the information that we do have. You defined god as a being capable of doing things based on will alone in a previous post and that so far we have no evidence that this exists. We are only on our one small planet and a few satellites (several of which we created). Evidence of this type of being could exist somewhere in the vast unexplored universe (the Star Trek episode “Who Mourns for Adonis?” comes to mind). I doubt it, but I am opened to it.
    .
    The conclusion that god is “not guilty” of existing that Matt has expressed in previous episodes is based on our lack of evidence for the existence of such a being and is a statement of belief. That is not to say that this statement of belief maps to a reality. The conclusion of said belief is based on our current evidence which we recognize is FAR from complete or ideal and is opened to new information which can change that conclusion. Furthermore, we don’t need to look at ALL evidence to come to that conclusion (though it IS a good idea). The conclusion is what is being explicitly considered in the courtroom.
    .
    Mind you Matt has suggested that there is evidence for god being “innocent” of existing in a few episodes (the context suggests that he is using the meaning as they would be in legal vernacular rather than common vernacular).
    .
    Does this make sense?

  332. corwyn says

    @EL:

    In the world where I come from, for a simple well-formed boolean proposition, either it’s true or it’s false. There is no third option.

    From this it is easy to show that you don’t live in the real world. 🙂
    “This coin will land heads up when I flip it tomorrow.” is a well-formed boolean proposition. It is currently neither true nor false.

  333. corwyn says

    @ frank:

    So a courtroom is NOT a place of pure Bayesian reasoning. It uses some ELEMENTS of Bayesian reasoning. A BIG part of the reason why it is not a place of pure Bayesian reasoning is the recognition that you won’t have all of the relevant evidence and there will be uncertainties. Even if you have overwhelming evidence you still do not have EVERYTHING. So what gets examined in these cases is an ESTIMATE of all potential hypotheses as suggested by the evidence presented.

    This is just wrong[1]. It is a complete misunderstanding about bayesian reasoning that it requires all the relevant evidence. It is exactly the opposite in fact, bayesian reasoning is the ONLY logically sound way of dealing with lack of complete evidence. Bayesian reasoning is simply the method for determining the probability of some hypothesis given some evidence (and some background information). That is what it does, and is provably the best method for doing so.

    Let’s say there is some blood found at the scene and (for the sake of argument) it has been proven to be the killer’s blood. If Alice has the same blood type, that is evidence that Alice committed the crime, yes? But not very strong evidence since many people share the same blood type, correct? So exactly how much weight should be given to that evidence? Bayes theorem answers that last question. A jury with a confidence in Alice’s guilt of X, should modify that confidence by a precise amount given by Baye’s equation when presented with this new evidence.

    [1] Actually courtrooms do not use baye’s theorem, but this is because they are terrible determiners of actual probabilities. (Well to be fair, also because doing the math is tremendously difficult) They certainly should.
    See: http://understandinguncertainty.org/court-appeal-bans-bayesian-probability-and-sherlock-holmes

    @EL

    Then, the jury should sum up all events that qualify as “Alice is actually guilty”, and if that number exceeds the reasonable doubts threshold, which maybe is *99%*, then the jury should deliver the legal verdict “guilty”. Otherwise, the jury should deliver the legal verdict “not guilty”.

    You wish! The actual number appears to be around 75-90%.

  334. corwyn says

    @EL:

    Do you think there’s a noteworthy difference between the words “foo” and “bar” as I have just defined them, and the everyday normal concepts of “guilt” and “innocent”?

    Yes.

  335. frankgturner says

    @ EL Still continued
    Also, I was pointing out that YECs tend not to understand evolution. Heck many don’t understand their own bible and have never read it. They need to “be right” ad view debate as a case of “winning” and “loosing” and see people as “wrong headed.” Many make no effort to try to understand evolution by reading about biology from a reputable source (some go to their pastors or people like WLC as authorities on the matter). From what I have seen they lack self esteem and are scared to be “wrong.” You don’t know how many of them I have met who have objections to Darwinian evolution who have never actually read “On the Origin of Species.” (I tell a story about one whom I motivated to actually read it, though I had to explain a lot as his understanding of biology was pretty limited).
    .
    I would think that you moved beyond that and would not have this “I must be right and Matt must be wrong” attitude. You did not seem to, particularly when you made the comments about telling people “how they are wrong, why they are wrong,” etc. That seems more like someone who needs to be right and beat down others. That seems just like YECs and apologists that I have spoken with who do that. It displayed a fairly closed minded attitude. I am hoping that you are getting past this.
    .
    I have explained to many YECs that such an attitude is not a great way to attract people. That repels many individuals and I have said to many of them that it is a reason why not to want to associate with them, even if I agreed with their beliefs.
    .
    I would hope that we are not only more appealing in a political sense (though this is not explicitly necessary) but that we convey a more opened minded attitude. That also comes with one’s mentality. If you are just trying to “be right” and to “win” rather than reach a level of understanding, then your drive to study what people are talking about can be affected. It took prompting for you to familiarize yourself with how legal terminology is used rather than it occurring to you to do it yourself. People like William lane Craig want to “be right” and “win” and make little to no effort to try to understand where people are coming from, Hence why he does not seem to see the need to study the hard sciences like cosmology and biology. He thinks arrogantly that he already has an understanding of said topics despite the obvious indication that he doesn’t as his usage of words reflects common parlance rather than the meaning used in said professional fields. (If you want someone who has made comments in opposition to Bayesian reasoning, you can have a field day with some of WLCs shit).
    .
    I am not saying that it is certain, but maintaining an attitude of trying to “reach a level of understanding,” in my personal experience, leads to an individual being prone to examining ideas for themselves wihtout prompting.

  336. frankgturner says

    @ corwyn #354
    What I was getting at is that EL has been presenting Bayesian reasoning this way, as though it were applicable because the situation was perfectly ideal.
    .
    I was waiting for him to point out the error but you beat him to it. (I wanted him to say it so that I could point out that in a legal setting using pure Bayesian reasoning is not such a good idea).
    .
    Though from what I have seen some Bayesian logic does go into legal affairs.

  337. EnlightenmentLiberal says

    @corwyn in 354
    Thank you for saying Frank got that bit wrong w.r.t. Bayesian reasoning. Maybe he’ll listen to you. He seemingly won’t listen to me when I tell him he doesn’t know what he’s talking about.

    You wish! The actual number appears to be around 75-90%.

    I do wish. To be fair, I still have very little idea what the proper threshold should be for “beyond reasonable doubt”. The numbers you cite do seem pretty low. The 75% number seems especially horrifying as the standard for criminal conviction.

    @Frank

    I was waiting for him to point out the error but you beat him to it. (I wanted him to say it so that I could point out that in a legal setting using pure Bayesian reasoning is not such a good idea).

    So, let me get this right. You made an argument which you knew was mistaken, as some sort of test or bait.

    Have you been regularly doing this kind of bait where you post something incorrect to see if I can catch it? That might be what’s feeding my impression that you don’t know what you’re talking about.

    Doing this kind of shenanigan just makes the conversation incredibly more difficult. Please don’t do this. Please say what you mean. Please don’t bait me. Please don’t “test” me. This kind of baiting and testing is a form of not arguing and discussing in good faith. Please knock it off.

    I actually think that a jury in a criminal court is probably a close to ideal setting to start formally using Bayes equation.

    Some minor odds and ends for Frank:

    Does this make sense?

    Most of it, with the following exceptions.

    A BIG part of the reason why it is not a place of pure Bayesian reasoning is the recognition that you won’t have all of the relevant evidence and there will be uncertainties.

    As corwyn said, this makes absolutely no sense. Either you are trolling me, testing me, baiting me, or you don’t know a damn thing about Bayesian reasoning. I’m still not sure which. The exact opposite is true. The recognition that there is uncertainties and lack of information is precisely what Bayesian reasoning is supposed to model and handle.

    THAT is a big part about why the analogy is used, i.e.: is god “guilty” of existing vs “not guilty” of existing based on our definition of god at the time and given the information that we do have.

    Yes, but an equally big part is this. The nature of the criminal justice system is that we would rather 10 guilty people go than convict 1 innocent person (approx). That’s why the jury delivers the verdicts “guilty” and “not guilty”. We are interested in whether:
    * the jury has a high degree of confidence that the person is guilty, or
    * the jury lacks a high degree of confidence that the person is guilty.

    A properly thinking jury has exactly 1 number between 0% and 100% that describes their degree of confidence that the person is guilty of the particular crime, and the statistical complement of that number is the confidence that the person is innocent (or positively not-guilty by reason of criminal insanity, or positively not-guilty by reason of justified self defense, etc.).

    100% = P(guilty) + P(innocent) + P(criminal insanity) + P(justifiable self defense) + etc.

    In other words:

    100% = estimation that he’s guilty + estimation that he’s innocent + estimation that he’s not-guilty by reason of criminal insanity + estimation that he’s not-guilty by reason of self defense + etc.

    Again, we place a very high bar to ensure that we only rarely convict innocent people at the expense of letting guilty people go unconvicted. That’s why the question to the jury is not “do you think that it’s more likely than he’s guilty than innocent?”, which is a 50% threshold. Rather, the question is “do you think that it’s likely that he’s guilty beyond a reasonable doubt?”, which is something more like a 90% or 99% threshold.

    I agree that the whole “guilty” “not guilty” verdicts thing is a thing because of incomplete information, however IMHO that’s the non-interesting part of the story. I just gave the interesting part of the story. That’s why it’s “guilty verdict / not guilty verdict” instead of “innocent verdict / guilty verdict”.

    The conclusion that god is “not guilty” of existing that Matt has expressed in previous episodes is based on our lack of evidence for the existence of such a being and is a statement of belief. That is not to say that this statement of belief maps to a reality.

    There is a difference between:
    * I am not convinced that a god exists.
    * I am convinced that no god exists.
    In your rather peculiar language, “god is ‘not guilty’ of existing” is identical to and interchangeable with “I am not convinced that a god exists” (which also happens to be the definition of an atheist).

    What I suspect people had the impression of here was that you insisted that you MUST demonstrate, in no uncertain terms, the “guilt” of another party to arrive at the conclusion that the accused party was “not guilty.”

    And I have absolutely no idea why anyone would think that. I’ve reread the entire thread several times now, and I still believe that I was more than clear for almost the entirety.

    Mind you Matt has suggested that there is evidence for god being “innocent” of existing in a few episodes (the context suggests that he is using the meaning as they would be in legal vernacular rather than common vernacular).

    I don’t know what you mean. I don’t know about the difference between the vernacular “that person is innocent” and the legal technical “that person is innocent”. Could you please explain the difference?

    Again, there is still a colossal miscommunication going on, because we still seem to have completely different understandings of what Matt has said in this very thread, relevant quote reproduced here:

    EL: You are wrong when you say that we can examine the question “is there a god?” without also examining the question “is it true that there are no gods?”.

    Matt: I’m sorry, but you’re just wrong in saying this…in part, because we’re NOT examining a question, we’re examining CLAIMS. The claim “some god exists” can be examined even if the claim “no god exists” cannot.

    Frank: I would think that you moved beyond that and would not have this “I must be right and Matt must be wrong” attitude.

    Never. Not until someone points out my error, or how I’m misinterpreting what Matt has said, or gives me something along those lines that points in the direction that I made an error or a misinterpretation of what Matt has said. I haven’t seen anyone even come close yet.

  338. EnlightenmentLiberal says

    Again, I know it’s been done a million times, but I would love for someone to know point out how i might be misunderstanding Matt in post 24.

    Here I go again:

    EL: You are wrong when you say that we can examine the question “is there a god?” without also examining the question “is it true that there are no gods?”.

    Matt: I’m sorry, but you’re just wrong in saying this…in part, because we’re NOT examining a question, we’re examining CLAIMS. The claim “some god exists” can be examined even if the claim “no god exists” cannot.

    Wrong. If you cannot examine competing claim “no gods exist”, then you also cannot examine the claim “some god exists”. Matt’s position violates the falsifiability criterion of science, and it violates proper Bayesian reasoning, and it violates the principles of critical thinking. Critical thinking is to think in a way that is critical of your own beliefs. In particular, critical thinking involves asking yourself “Am I wrong? How can I be wrong? How would I know if I’m wrong?”. In this case, Matt’ has determined that he can answer positively to a question “does some god exist”, without being able to answer the critical thinking questions “How can I be wrong? How would I know if I’m wrong?”.

    The claim “this jar has an even number of gumballs” can be accepted or rejected, without ever considering the contrary proposition “this jar has an odd number of gumballs”.

    This is a bad example, because there are only two outcomes to choose from, and it’s hard to separate intuition from the formal reasoning. The problem is that any data that we intuitively think is evidence for even-ness also happens to not be evidence for odd-ness. That determination is not so easy to make in more complicated scenarios. It gets back to what I said earlier: The only way that you can determine if a piece of data is evidence for a proposition is to examine all of the (plausible) competing claims, and determine that the piece of data is more likely to exist / be found on one proposition compared to all of the rest. The only way to properly examine a claim like “the number of gumballs is even” is to determine what evidence you have for that proposition, and as just explained that requires trying to fit the known data to all competing claims.

    Further, depending on semantics, one might also have evidence in favor of oddness, and you need to take that into account when examining the claim of even-ness. For example, I might have a sworn signed statement that the number is actually odd. I need to take that into account when examining the claim of even-ness. Otherwise I’m not practicing proper reasoning.

    I can accept one or none of the possibilities. There may not be sufficient information to determine whether the number is even or odd

    Correct.

    – but the person who asserts that it IS even (or odd) has adopted a burden of proof. If you tell me the number is even, I can assess your evidence FOR that proposition, without ever being required to consider evidence for the contrary (which there may not be).

    I agree. However, this is a cultural contrivance for fairness to ensure that we don’t waste each other’s time. It’s not a basic fact of reasoning. The truth of the matter does not depend on if someone speaks it, or who speaks it. Further, our proper estimation of the matter should not depend on if someone speaks a claim, or who speaks a claim.

    Rather, this particular notion of the burden of proof is a cultural convention, just like spontaneously forming a line for order of check-out at a grocery store. We recognize that this particular rule governing interpersonal conduct is fairer than some alternatives. If someone makes a random claim, it’s not my job to find evidence for that claim. That’s all the burden of proof means.

    For example, if someone makes some claim, and if I choose to honestly engage with that person, there is an onus on me in the conversation to engage honestly, to take into account all of my known evidence and data, etc., so it may be that if someone makes a claim, I might already know sufficient data to confirm that claim, and the honest and reasonable person that is me is compelled to agree with that person and the claim in spite of the lack of additional evidence provided by the other person.

    PS: I can choose to not engage with the person without violating my principles of honesty and reasonableness.

    I can accept your claim that the number is even, or reject it.

    Correct.

    Rejecting your claim does not mean that I’m convinced the number is odd, it doesn’t mean that I think it’s more likely that the number is odd, it doesn’t mean that I’m “leaning” toward the chance that it’s odd […]

    Correct.

    and it doesn’t mean that I’m 10%, 20%, 40%, 60% in favor of it being odd.

    Wrong.

    If someone rejects my claim of even, that means that they are not convinced that the claim of even is true. That means that their estimation of the even claim is less than say 90%. Again, that necessarily means that their estimation of the odd claim should be 10% or more because a rational actor’s estimations must obey:
    100% = P(odd) + P(even)
    (give or take the problem of partial or incomplete gumballs).

    So, by telling me something about your position on the claim “the number of gumballs is even”, I also learn something about your position on the claim “the number of gumballs is odd”. If you tell me that you’re not convinced that it’s even, I learn something about your estimation that the number is odd.

    I did not say that I learn that you believe it’s odd, or leaning that it’s odd. However, I did learn that you don’t have an estimation of odd which is 10% or less.

    Again, if you did have an estimation of 10% or less of oddness, that means that you have a strong confidence that the number is not odd, e.g. you have a strong confidence that it is false that the number is odd. If you’re a rational logical agent, that necessarily means that you must hold a 90% or more estimate that the number is even, which means – by my arbitrarily chosen cut-off of 90% – that you believe that the number is even. Reductio ad absurdum: if you do not accept that the number of gumballs is even, it necessarily follows that your estimate of oddness is 10% or higher.

    To put it another way, it’s logically inconsistent to hold estimations of 20% oddness and 20% evenness. It’s logically inconsistent to hold estimations of 80% oddness and 80% evenness. Whatever estimates you have must obey:
    100% = P(odd) + P(even)
    Otherwise you’re doing it wrong.

    (Again, assuming a whole number of gumballs. I can repeat this same analysis taking that possibility into account, but it doesn’t change the overall principle.)

  339. corwyn says

    @ 358 EL:

    in part, because we’re NOT examining a question, we’re examining CLAIMS. The claim “some god exists” can be examined even if the claim “no god exists” cannot.

    Let’s see if I can rephrase Matt’s statement both correctly, and so the distinction is clearer.

    The claim that a god exists can be translated into ‘there is sufficient evidence to provide a probability of a god’s existence to proper bayesian analysis as above the credibility threshold’. Yes? If there was no evidence *at all* presented, one would properly determine that the probability was identical to the prior. Yes? So unless one is so adle-brained as to start a priori assuming something’s existence (particularly something as extraordinary as a god; even most christians will concede that there are at most 3 in the entire universe), the prior should not exceed the credibility threshold. Yes? Therefore a claim backed by no evidence can be rejected with no other requirements.

    This is precisely what we would expect in a courtroom as well. If the prosecution rests before opening remarks, the judge will properly just dismiss the case, without waiting to hear from the defense.

    This is what I think Matt is trying to say.

  340. frankgturner says

    @EL

    Frank: I would think that you moved beyond that and would not have this “I must be right and Matt must be wrong” attitude.
    Never. Not until someone points out my error, or how I’m misinterpreting what Matt has said, or gives me something along those lines that points in the direction that I made an error or a misinterpretation of what Matt has said.

    The reason that I bring this up is that this is an attitude many apologists and YECs and evangelists have. They care not about reaching an understanding or solving a problem, they just want to “win” the argument, so someone, their opponent, needs to be “wrong.” That attitude is not productive and often leads to people not trying to make alternate considerations for what the other individual is trying to say, what their reasoning is and the like. When I made the comment about “supernatural” and tried to relate the idea of how the word “supernatural” was being interpreted in that context, it took a LOT for you to consider what I was talking about. You jumped to what I perceived to be an attack and made accusations of being intellectual dishonesty. A no point did I say that in no uncertain terms that supernatural and non existent were perfectly identical, merely that in the context I was speaking they had similar enough meanings to be used interchangeably. If you had flat out asked me I would have told you that I did not intend for them to be viewed as identical, but you seemed to focused on being “right” about you assessment of the situation to consider asking (much like YECs are too focused on being “right” about creationism to make an effort to understand evolution).
    .
    Other people in here who were not so determined to demonstrate that I was “wrong” in equating (or rather synonymizing) the terms in context recognized that. You didn’t. One might suspect that even if you saw evidence that this were not the case you would have ignored it given that it did not agree with the feeling you so passionately had about it. You should recognize that, it is called “confirmation bias” and apologists and YECs and evangelists do that a lot too.
    .
    I am not saying that you always do that, but maintaining a model of needing to “win” and insisting that you won’t back down unless it is absolutely demonstrated that you were in error will lead you to that type of closed mindedness we see in apologists, YECs, and evangelists. I still think thats why it took you so long to look up the differences between how the words “innocent” and “guilty” are used in common vernacular vs in legal terms. (FYI, the wikipedia articles you linked are exactly the ones I would have sent you to had you not taken the suggestion and done it yourself, I should know I had them bookmarked). Who knows though, maybe you just can’t let go of that “win-loose” model in your head.
    .
    And that is part of why I thought of baiting you (and did so) at least twice in here. I started to figure that the only way to get you to consider alternate ways of thinking was to make you look like an idiot (you did a good enough job of making yourself look like an asshole). You behave in a way that suggests that you are closed to alternate evidence. It almost sounded like you would defend the common definitions of “innocent” and “guilty” as being “more correct” than what we had been trying to tell you (how they are used in legal terminology) rather than admit you were in error (particularly given that the context was HIGHLY suggestive that the terms were being used in proper legal parlance). You are thinking about what they mean though and that is good.
    .
    I will point out that in several places I misapplied Bayesian reasoning. I make mistakes about context too. We all do because we are human. (And sometimes I was a dolt too, not seeing what was being suggested by context).
    .
    I don’t disagree with corwyn’s interpretation of what Matt might have been saying. That is essentially the point of a phrase Matt says a lot, “a proposal that is made without evidence can be dismissed without evidence.” You might like that as it is nice and Boolean.
    .
    There is a lot of speculation in what follows so don’t think that I am stating any of this as solid. This is merely suggested by several thoughts. (So don’t “get your panties in a bunch” if you disagree with this, I am not solid on it myself). I might point out that Matt is talking about is with regard to belief. That might be where he and Carrier differ in principle. That is suggested by the fact that they are using the word “claim,” which in mathematical context is a bit different from how it is used in the common vernacular and particularly how it is used in legal circumstances. In a courtroom a “claim” is like a formal charge in that it is a statement of belief. In that regard making a “claim” about the existence of god is not just “god exists” or “god does not exist,” it is “we believe god exists” or “we do not believe god exists.” That is a bit different from “God exists” vs “God does not exist.” One is competing statements of belief, the other is competing statements about potential fact. In one case we are examining what people think and in the other how what they think maps to reality.
    .
    That is similar to why I separated out the action (stabbing) and whether it occurred or not from whether that action should be considered a crime or not. For situations in which we are discussing action I agree with Carrier’s statement that you cannot examine the claim (which in his case is NOT a statement of belief like in the courtroom) without examining all competing claims. Someone either did something or did not do it (at least with respect to a limited time period). In that respect the word claim is not a statement of belief, but a statement of how the belief maps to reality.
    .
    Belief is a much fuzzier question. The “event space” (if you want to call it that) is not strictly limited to “A person either believes something or they do not.” And in many ways you can make a conscious rational effort to examine why an individual does believe something and ignore evidence about why they do not believe the opposite (though it is not necessarily a good idea). FYI, I am not saying that the term “belief” is bullshit like the word “supernatural,” just that in many cases it is not strictly Boolean.
    .
    One of the big problems is the ambiguity of the English language. Still corwyn’s explanation makes a lot of sense given the Google definition of the word “claim”

    claim/klām/
    verb
    state or assert that something is the case, typically without providing evidence or proof.
    noun
    an assertion of the truth of something, typically one that is disputed or in doubt.
    a demand or request for something considered one’s due.

  341. EnlightenmentLiberal says

    @corwyn
    Ok. I’m generally confused why Matt would choose the word “examined” instead of other words like “demonstrated”, “asserted”, “substantiated”, “defended”, “justified”, etc. “Cannot examine” has a stronger meaning.

    My reading is still probably the intended meaning based on these considerations.

    Until recently, (and maybe even now), Matt has promoted unfalsifiable magical thinking by saying that there is a natural world and it’s impossible for science to demonstrate the existence of supernatural stuff or supernatural causation. It’s the same flaw of reasoning which is in the above posts of Matt. It’s the flawed idea where Matt thinks that you can address, examine, analyze, and be able to accept or reject a claim without examining counter-claims.

    Further, it also fits nicely with how the hosts of the show talk about the burden of proof and the definition of atheism. IMHO, the hosts of the show misuse the burden of proof when discussing atheism, and also in other contexts as well, such as criminal trials and guilty vs innocent. The hosts have often said that the atheist does not need to present a positive defense of atheism when arguing with a Christian, and it is sufficient to merely attack the Christian’s arguments for god. That is wrong. I only realized that recently.

    As Sherlock Holmes once said, “once you eliminate the impossible, whatever remains, however improbable, must be the truth”. No amount of attacking the Christian’s argument will do any good if the Christian doesn’t currently know of an alternative explanation. Matt would say that several common Christian arguments for the existence of god are fallacious appeals to ignorance. What I have come to realize is that the form of the argument is not fallacious. It’s the particular premises which make it fallacious. Sherlock Holmes is entirely right in that quote. Not all arguments of that form are fallacious.

    The reason why several of those common Christian arguments for god are fallacious is because they have an incomplete listing of the epistemically-possible space of how the world might really be (the event space). That’s where an atheist needs to at least present alternative explanations of the available evidence – plausible, not demonstrated, not “more likely true than false”, just plausible. You need to show that it’s at least plausible in order for the Christian to back down from their claims. If the Christian is reasoning correctly, no amount of attacking the Christian’s argument will do a lick of good if the Christian believes that the alternatives are even more impossible. That’s what Sherlock Holmes tells us with “if you eliminate the impossible, whatever remains, however improbable, must be the truth”. That’s what Bayesian reasoning teaches us.

    I know this is radical claim in this place. I know this wildly contradicts the philosophy and epistemology of the hosts. However, it’s as simple as I’m right and they’re wrong, and people need to learn more about Bayesian reasoning.

    @Frank
    You are wrongly confusing a tenacity for the truth for a desire to win the argument at all costs. You’ve made this wrong characterization of me in about 10 posts now. Make it all you want. I reject it completely. Your tone trolling is completely and utterly ineffective.

    “Tone trolling”. Definition: Attacking the tone or style of someone’s arguments in contrast to attacking the substance of the argument. Often used as a way to attack the argument without engaging with the substance of the argument. See: ad hominem.

    Also, I never backed off that claim of intellectual dishonesty. Your communication skills are poor, and you are either continuing to dishonestly weasel, or dishonestly move the goalposts, or you are in denial about what you actually said. I invite you to go back and read it.

    I suggest you go back and double check your facts and understanding. Just off the top of my head, you implied / stated that you have been here longer than I – almost certainly a mistake and bullshit you pulled out of your ass without even bothering to check it. You confused me with Adamah, even though I was engaging in many of those conversations with Adamah under this name. And that confusion with me with Adamah is probably now coloring your perceptions of me. In addition, it’s painfully obvious that you don’t know what you’re talking about w.r.t. Bayesian reasoning, and I think some Dunning–Kruger thing is going on now, which is making you react negatively to me when I flatly tell you that you don’t know what you’re talking about. I suggest that you shut up and read a book, such as Proving History, because I am sick and tired of your tone trolling.

    I will point out that in several places I misapplied Bayesian reasoning. I make mistakes about context too. We all do because we are human. (And sometimes I was a dolt too, not seeing what was being suggested by context).

    There’s occasional mistakes, and there’s repeated fundamental mistakes. You have repeatedly misapplied Bayesian reasoning in crucial, fundamental ways. You have made repeated mistakes in such a way that it’s highly unlikely that you actually know what you’re talking about. The explanation “it’s a one-off mistake” does a very bad job of explaining the available data.

    The “event space” (if you want to call it that) is not strictly limited to “A person either believes something or they do not.”

    No, actually, I don’t want to call it that. In fact, I think calling it that is highly confusing, and again this leads me to the conclusion that you don’t know what you’re talking about. Again, please read. In the courtroom example, the event space includes;
    * person committed the crime with intent
    * the person did not perform the criminal physical actions
    * the person committed the criminal physical actions but without intent
    * etc.

    There’s this other thing called the belief space. It’s not the event space. I give it a different name because it’s a different thing. The belief space includes:
    * I believe the person committed the crime with intent, with high confidence
    * I believe the person committed the crime with intent, with lower confidence
    * I have a hunch that the person committed the crime, but I’m not sure
    * I don’t know if the person committed the crime / I don’t know / I have no opinion
    * I believe the person did not commit the physical criminal actions, with high confidence
    * I believe the person did not commit the physical criminal actions, with lower confidence
    * I have a hunch that the person did not commit the physical criminal actions
    * etc.

    FYI, I am not saying that the term “belief” is bullshit like the word “supernatural,” just that in many cases it is not strictly Boolean.

    Beliefs are never binary. That’s the whole point.

    If you think that I have ever stated that beliefs are binary in the sense that you either believe or you don’t with no middle ground or degrees of confidence, then you’re projecting. Your beliefs that I’m not reading for comprehension and I’m playing just to win are instead an accurate description of you, and the Dunning–Kruger effect would be amazingly strong with you.

  342. EnlightenmentLiberal says

    To continue:

    You need to show that it’s at least plausible in order for the Christian to back down from their claims. If the Christian is reasoning correctly, no amount of attacking the Christian’s argument will do a lick of good if the Christian believes that the alternatives are even more impossible. That’s what Sherlock Holmes tells us with “if you eliminate the impossible, whatever remains, however improbable, must be the truth”. That’s what Bayesian reasoning teaches us.

    This is the same point we correctly make when Christians argue against evolution. No matter how much one argues against evolution, creationism does not win by default. Rather, Christians also need to present positive evidence in favor of creationism.

    Using the same principles of reasoning, no amount of tearing down a Christian’s arguments should convince them that they’re wrong if you don’t at least present alternatives and defend the alternatives as at least plausible.

    We’re always examining competing claims.

    In the case of evolution vs creationism, the wealth of evidence is so strong for evolution and so strongly against creationism that no amount of tearing down evolution will do, because there’s so many other options that are not creationist and because we have a wealth of evidence against creationism.

    In the case of many Christians, they are convinced that “no god” alternatives are impossible, and so they are applying correct reasoning when they continue to cling to their god claim no matter how much we tear it down. The only way we can make progress is also tear away at their false notions that “no god” alternative are impossible.

  343. Monocle Smile says

    @EL

    Ok. I’m generally confused why Matt would choose the word “examined” instead of other words like “demonstrated”, “asserted”, “substantiated”, “defended”, “justified”, etc. “Cannot examine” has a stronger meaning.

    YES. I’ve actually wanted to bring this up for a bit. Depending on the clarification of what Matt means by “examined,” I might throw lots of my defense of him overboard and agree with you on this point.

    In the case of many Christians, they are convinced that “no god” alternatives are impossible, and so they are applying correct reasoning when they continue to cling to their god claim no matter how much we tear it down

    It’s like a train that hits a railroad switch and goes down the wrong path. The train runs just fine down that path, but it took a wrong turn early.
    However, I’m not totally in line with the implication. This assumes intellectual honesty and the genuine desire to hold true beliefs regardless of what those beliefs might entail. I’ve become quite the cynic in this department; I feel that most christians are intellectually dishonest at some level (or indoctrinated to operate as if they are).

  344. frankgturner says

    @ EL
    Well if it isn’t the pot calling the kettle black.
    Regarding believe space, I was speculating. I even stated it. And reguarding winning at any cost, you don’t make an effort to look many things up until prompted. That’s not solid proof, just evidence in favor of that type of thinking. Which I have even stated several times is characteristic of an individual trying to win at any cost, not hard proof
    .
    My tone trolling of you as not having tenacity for the facts was stated as that, characteristic. You don’t read carefully do you? I guess that I should expect that, you’ve accursed me of things that Matt has said, MS of things that John has said, eall of us of things that no one has said etc. Then again at times my communication skills are poor. So are yours, you have shown evidence of getting too emotionally wrapped up in your arguments to look at them carefully.

  345. frankgturner says

    @ MS
    Human beings are highly emotional creatures. Indoctrination has a characteristic of burying people in a deep emotional desire to feel a certain way. Their belief in their God makes them feel good. They don’t want the good feeling to go away so they “need” to believe (James Randi talks about this).
    .
    I read somewhere that it is like talking to drug addicts. Even entertaining the idea that their beliefs might be false could make them terrified that the good feeling will go away and that they won’t have it. And they look for enablers too (other addicts).

  346. frankgturner says

    @ MS continued
    One of the things I am doing that others seem to miss is that I am constantly considering (one might even say “examining”) competing word meanings. I think that I picked that up a long time ago when I learned about context clues (which used to drive me bonkers but I now see is actually a good thing).
    .
    I did that when Matt used the word “supernatural” with YEC’s and I would tend to do that with the word “examine,” but I do agree that there are better terms. I would suspect that if pressed on the concept of “mathematical” examination (which from context is what I think Carrier is talking about) Matt would back off and choose a different verb. I could be wrong though.
    .
    That is something that comes up in courtrooms. In many cases they only consider something “examined” if a formal statement was made about it, regardless of if implications are strong (occassionally though implications are so strong that they don’t need to be stated). So (to pull an analogy from earlier) in that case if the probability of the gumballs is odd is 60% the court does not consider the probability of the gum balls being even to be 40% unless someone formally makes that statement (it sounds stupid as the implication is so obvious that it should not need to be stated but courts are weird like that).
    .
    As a juror my father saw that. After obvious implications (in favor of the defense, the fact that the prosecution would even ask such questions was ridiculous as it made the defense’s case) the defending attorney got up and said, “just to be clear, because your mother was in need of medical treatment and could die if not gotten to a hospital, you drive her to the emergency room despite having a suspended driver’s license, is this correct?”. Some judges will sop that stuff due to redundancy and some won’t.
    .
    That’s a big part of why I talked about implied vs explicit consideration and why the court is not strictly Bayesian but used elements of Bayesian reasoning.

  347. frankgturner says

    @ MS and corwyn
    Come to think of it, it does make a LOT of sense that things are not classified as “considered” and “examined” unless formally stated in the courtroom even if it is painfully obvious that they are implied and why the percent confidence for guilt would be so low (75%) to reach a “guilty” verdict. The average juror winds up being of low education level.
    .http://thebilzerianreport.com/jury-of-our-peers/

  348. corwyn says

    @EL:

    That’s what Sherlock Holmes tells us with “if you eliminate the impossible, whatever remains, however improbable, must be the truth”. That’s what Bayesian reasoning teaches us.

    I sure hope not. It certainly isn’t what Bayesian reasoning teaches *me*. In Bayesian reasoning there is no such thing as ‘impossible’. That would imply a probability of 0, which would in turn require *infinite* evidence (do the math, see for yourself). Bayesian reasoning requires that *every* hypothesis is *always* in play, just that its probability gets extremely low.

    Here you are, making emphatic statements about Bayesian reasoning again, and again being wrong. Please stop doing that.

  349. frankgturner says

    @ corwyn
    To be fair and extend a courtesy to EL that I doubt he would return (giving him the benefit of the doubt). He may have intended that some hypotheses be considered statistically insignificant so as to be ignored with the statement despite it not being implied as it is often associated. Given the strong association I would not consider that “weaseling” or “moving the goal posts.”

  350. EnlightenmentLiberal says

    @corwyn
    I see that you’re still on this unreasonable crusade against me. It’s really hurting your reading comprehension. I really think that if you just calmed down a little and dropped this unreasonable grudge against me, then we could have a much more productive conversation.

    I never advocated nor defended having 0% confidence in a proposition. I never advocated nor defended believing something is impossible to absolute 100% confidence. Rather, I was explaining the fallacious beliefs of some specific hypothetical Christians. The position of the Christians fallacious – I explicitly called that fallacious. I’m not defending fallacious reasoning.

    Even then, I never clearly stated that the Christians had a literal absolute 0% confidence or prior. Rather, I merely said “belief that it’s impossible”. You’re doing the mistake of inferring absolute confidence from everyday English where no absolute confidence is implied. It would be a similar mistake to hear “I know X is true” and reply “Aha! You have absolute certainty!”. I even used the phrase “more impossible”, which should be another dead give-away that I’m not talking about actual infallible absolute confidence.

    Interpreting “impossible” as “absolute 0% confidence” in the Sherlock Holmes quote is unreasonable, especially for you corwyn because you should know better. It’s just as unreasonable as interpreting “I know” to be a statement of infallible absolute confidence. I know that you know that what Sherlock Holmes is saying is that if one scenario is ridiculous, but if all of the other scenarios are even more ridiculous, then the scenario which is far less ridiculous than the rest is very probably the right answer.

    I really don’t know what to do with you.

    @Frank, if you want to see someone who is trying to win at all costs, it’s corwyn, not me.

    @MS and Frank
    Matt might have the right ideas on what constitutes proper reasoning in this case. I’m dubious of that. Frank mentions the intellectual equivalent of the principle of charity. The principle of charity strongly applies in one-on-one conversations like what we’re having. However, the principle of charity only weakly applies to my complaint against Matt. Matt is a public speaker and is teaching / advocating a certain style of reasoning. In that situation, the principle of charity is a much weaker thing. Regardless of whether Matt actually has the right ideas concerning proper reasoning, Matt chose the wrong word “examine”, causing him to communicate the wrong idea. The onus is on Matt to choose the right words, and the error is Matt’s.

    From Merriam Webster:

    Examine
    1
    a : to inspect closely
    b : to test the condition of
    c : to inquire into carefully : investigate
    2
    a : to interrogate closely
    b : to test by questioning in order to determine progress, fitness, or knowledge

    Further, Matt seems to use the following words in a mostly interchangeable way in the original context:
    * examine
    * consider
    * assess

    Frank, are there similar obscure, esoteric legal meanings for “assess” and “consider” as there is for “examine”?

    @Frank in particular

    It seems that Frank is again operating under some highly specific and esoteric usages of the English language. I was going to ask him to explain himself, but he does so pre-emptively here:

    That is something that comes up in courtrooms. In many cases they only consider something “examined” if a formal statement was made about it, regardless of if implications are strong (occassionally though implications are so strong that they don’t need to be stated). So (to pull an analogy from earlier) in that case if the probability of the gumballs is odd is 60% the court does not consider the probability of the gum balls being even to be 40% unless someone formally makes that statement (it sounds stupid as the implication is so obvious that it should not need to be stated but courts are weird like that).

    That’s a big part of why I talked about implied vs explicit consideration and why the court is not strictly Bayesian but used elements of Bayesian reasoning.

    I’m starting to understand where Frank is coming from now.

    A process of gathering evidence and presenting evidence to a jury cannot be Bayesian, and it cannot be “not Bayesian”. Rather, Bayesian describes the internal process by which the jury determines estimations of guilt, innocence, etc., which serves as the basis for determining which verdict to return.

    I’m highly confident that Matt is not using your peculiar esoteric definition of “examine” to mean (I hope I get this right) “something is examined iff a formal process of investigation has been made in a court of law, which ended in a formal conclusion” or something like that. At the time of the quote,

    I’m now beginning to understand what you mean by implicit vs explicit. However, I again think that you’re missing the point. Bayesian reasoning is what goes on in the heads of the jury, not in the methods and processes of presenting data to the jury. It may be that the methods and processes of presenting data to the jury could be changed to be better suited to a better informed Bayesian analysis, but it’s still improper to describe the methods and processes of the court as Bayesian nor not-Bayesian.

    As mentioned above, Matt seems to use the following words in a mostly interchangeable way in the original context:
    * examine
    * consider
    * assess

    Frank, are there similar obscure, esoteric legal meanings for “assess” and “consider” as there is for “examine”?

    @Frank

    And reguarding winning at any cost, you don’t make an effort to look many things up until prompted.

    For example, near the start of the thread transcribed several quotes of Matt myself, unprompted, in order to demonstrate my point. Even here, I’m copied the Merriam Webster definition of “examine”. I don’t know what you’re talking about. Could you please provide some specific examples where you feel that I should have looked something up when I did not?

  351. John Iacoletti says

    @EL, 347
    I was totally with you on this whole post until you got here:

    Do you agree with the reasoning of Andy’s fellow prisoners? Do you think that they did proper reasoning when they learned of evidence (really bad evidence, hearsay) that someone was committed the crime, and from that significantly lowered their estimation of the likelihood of Andy’s guilt?

    No, I don’t agree. You could have 100% indisputable proof that unnamed lover killed Andy’s wife. That doesn’t mean that Andy did or did not also kill his wife — unless you have other evidence that exactly one person killed her, but then that spoils the general rule you’re trying to make.

    #359:

    I can accept your claim that the number is even, or reject it.

    Correct.

    Rejecting your claim does not mean that I’m convinced the number is odd, it doesn’t mean that I think it’s more likely that the number is odd, it doesn’t mean that I’m “leaning” toward the chance that it’s odd […]

    Correct.

    and it doesn’t mean that I’m 10%, 20%, 40%, 60% in favor of it being odd.

    Wrong.

    No, you are wrong, and here’s why. In the absence of any valid evidence, I have 0% confidence in the claim “the number of gumballs is even”, and I also have 0% confidence in the claim “the number of gumballs is odd”. My 0% confidence in the first claim causes me to reject it. That does not affect or change my confidence in the second claim, which remains 0. Without evidence, I reject both claims.

  352. EnlightenmentLiberal says

    @John
    First question – how do you underline in posts? I looked for half a minute one time and couldn’t figure it out. (I didn’t try very hard.)

    Let’s start with this.

    No, you are wrong, and here’s why. In the absence of any valid evidence, I have 0% confidence in the claim “the number of gumballs is even”, and I also have 0% confidence in the claim “the number of gumballs is odd”. My 0% confidence in the first claim causes me to reject it. That does not affect or change my confidence in the second claim, which remains 0. Without evidence, I reject both claims.

    In my defense, I thought I explained this well. However, I could do better. Let me try again.

    This is a common stumbling block. It also has to do with possible disagreement – legitimate disagreement – over the meaning of the word “confidence”. I’m using the word “confidence” in a rather specific esoteric way. In order to avoid confusion, I’ll try to completely avoid using that word.

    Richard Carrier has noted that using odds instead of confidences is more intuitive to a lot of people. Let me try it like that.

    As I said, we always have incomplete information. Every decision we make is based on some degree of incomplete information. Every decision is the result of a (formal or informal) cost-benefit analysis. That’s what it means to be a rational agent. A cost-benefit analysis in this sense is this: For a particular action, we first determine the list of (epistemically) possible outcomes, then we determine the negatives and positives of each outcome, then we determine an estimation of the likelihood of each outcome. We repeat this for various actions, and then we do a value judgment comparison of the actions. Each action doesn’t have a single outcome, but a plethora, a spectrum of outcomes, where each outcome has an assigned estimation aka probability, and each outcome has an assigned negative and positive value.

    In order to be this rational agent, in order to do this cost benefit analysis, we need to create estimates for the (epistemically) possible outcomes of our actions.

    In a more general sense, we need to be able to estimate the likelihood of the truth of various propositions, and we need to do so based on incomplete information. That’s what Bayes equation helps you do.

    The first thing you need to accept is that you always have an estimation of the likelihood of a (well-defined) event. Perhaps it’s not a very good estimate because you haven’t taken time to gather data, and perhaps it’s not a very good estimate because you haven’t taken the time to formally list out the available data, analyze the data, and come to a proper conclusion. The answer “I don’t know” to the question “does the Christian god exist?” is fine for everyday conversations, but it’s an incomplete answer. You do have an estimation of how likely it is, and you need that estimation for going about your daily business. One of the reasons that Pascal’s Wager fails is because our proper estimation of the likelihood of the existence of the Christian god is near 0. If our estimation of the Christian god was really 50 50 odds, it may be entirely rational to take Pascal’s Wager (ignoring the plethora of other problems with it).

    Bayesian reasoning deals with only 1 kind of number: the actual real-world gambler odds of the truth of the proposition. Previously, I have used the words “confidence”, “probability”, “estimation”, and “(gambler) odds” to all refer to this same concept.

    Bayesian reasoning is not frequentist reasoning. In frequentist reasoning, one gives a likely bounds on the true value, a confidence interval, and a degree of confidence that this bounds is accurate. That’s two kinds of numbers, 1- confidence intervals, and 2- confidence levels. Bayesian reasoning is not that. Bayesian deals with only 1 kind of number – the gambler odds of the actual real-world truth of the proposition.

    Refresher: “X to Y” odds is by definition equal to “X / (X + Y)” as a probability fraction. Example: “2 to 3” odds is by definition equal to the probability fraction “2 / (2 + 3) = 2/5 = 0.4”.

    Refresher: A probability fraction like 0.4 is the same number as 40%. You should think of the percent sign “%” as a dimensionless unit that means “1 / 100”. Ex: “30% = (30) (1/100) = 0.3”.

    Absent any particular evidence that would favor “even gumballs” or “odd gumballs”, you as a good rational actor should have an estimation of the truth of the truth of the proposition “the number of gumballs is even”, and that estimation should be about 50 50 odds, aka 50% likelihood of truth. Similarly, without particular evidence, you should also have an estimation of the likelihood of the truth of the proposition “the number of gumballs is odd”, and that estimation should be 50 50 odds, aka 50% likelihood of truth.

    Again, it is a simple fact of being a good gambler, aka a rational actor, that for any set of complete and mutually exhaustive (epistemic) possibilities of how the real world actually is, the sum of the estimations of the likelihoods of truth should equal 1 aka 100%. For the case of gumballs, I might have the estimates:
    P(odd gumballs) = 50%
    P(even gumballs) = 50%
    100% = ∑P(X) = P(odd gumballs) + P(even gumballs)

    (Of course, it may be that the number of gumballs is neither even nor odd because maybe there’s a half of a gumball, but this can also be modeled under the same framework.)

    As a rational gambler, these numbers simply describe to you what kind of bets you would be willing to take. With 50 50 odds for even, you should take a bet where you win net $2 for even and lose net $1 for odd. That’s because the expected payout is positive.
    expected payout
    = (+2$)(50%) + (-1$)(50%)
    = (+2$)(.5) + (-1$)(.5)
    = 1$ – .5$
    = .5$

    Again, you already do this kind of bet every day of your life. Every action you take is this kind of bet because you’re always working on odds. Your cost benefit analysis that you do for every decision must take into account these kinds of odds that you have because of your incomplete information.

    (Another consideration is this: If someone offers you a bet like this, that offer is itself additional information. This may suggest that the person has additional information if they’re offering the bet, and they might be trying to scam you. A good rational actor would not ignore this additional information of human psychology, human motivations.)

    We can talk about different propositions, like “the number of gumballs is divisible by 4”. A good rational actor, absent any further information, should have the estimates:
    P(gumballs divisible by 4) = 25%
    P(gumballs not divisible by 4) = 75%
    100% = ∑P(X) = P(gumballs divisible by 4) + P(gumballs not divisible by 4)

    Again, this informs you what bet you would be willing to take. With these odds, it would be foolish to take a bet where you win net $2 for “divisible by 4” and lose net $1 for “not divisible by 4”. That’s because the expected payout is negative.
    expected payout
    = (+2$)(25%) + (-1$)(75%)
    = (2$)(.25) + (-1$)(.75)
    = .5$ – .75$
    = -.25$

    I can talk about the same thing for a jury trial. In my hypothetical state of pure ignorance, knowing nothing about jury trials, starting with 50 50 odds is prudent. Now, assume I learn that the defendant is one person from the population of the world, and that it’s likely that the number of people guilty for this crime is a very small number. That knowledge should immediately cause me to adjust my estimate from 50 50 odds to something like 1 to 6 billion odds. As further evidence comes in, I should further adjust my estimate of guilt.

    If I am convinced beyond a reasonable doubt that the person is guilty of the particular crime, then I should vote for the guilty verdict. If I am not convinced beyond a reasonable doubt that the person is guilty of the particular crime, then I should vote for the not-guilty verdict. In the Bayesian mindset, what it means to be convinced beyond a reasonable doubt is to have a particular estimate which is higher than some threshold like 1 to 100 odds, e.g. approx 99%. If you merely have 2 to 1 odds that you’re right, e.g. approx 66%, that’s not “convinced beyond a reasonable doubt”. There’s a 1 out of 3 chance that you’re wrong, and hopefully we can agree that “1 out of 3 chance that I’m wrong” is not “convinced beyond a reasonable doubt”.

    Similarly, what it means to have a belief that some proposition P is true is not “I believe it’s more likely to be true than false”. It’s not good enough to merely have 2 to 1 odds that P is true. You might have an estimation that the odds are 2 to 1, but that’s not enough to meet the usual burden of being convinced, and thus you don’t have the belief “P is true”. You need to have a higher estimate of truth in order to meet the requirements to say “I am convinced that P is true” or “I believe that P is true”. For example, perhaps 1 to 100 odds, e.g. approx 99% odds, may be an appropriate bar in order to declare “I am convinced that P is true” and “I believe that P is true”.

    John, I know that you’ve not been taught to think this way, and that’s why I’m trying to get you to think this way. The other way of thinking about things where you only analyze, consider, examine, assess, etc., the data in light of one proposition without analyzing, considering, examining, assessing, etc., competing propositions is bunk.

    No, I don’t agree. You could have 100% indisputable proof that unnamed lover killed Andy’s wife. That doesn’t mean that Andy did or did not also kill his wife — unless you have other evidence that exactly one person killed her, but then that spoils the general rule you’re trying to make.

    I do appreciate this point of logic, and in that point of logic you are entirely correct. You do need additional evidence that only 1 person committed the murder in order to make my conclusion.

    However, you do have that evidence. Part of your background knowledge should include the fact that a majority of murders are committed by a single person working alone and not by a group (citation needed). You don’t need evidence particular to the case to use this general fact about the statistical distribution of murderers. If we are convinced (not absolutely convinced, just reasonably convinced) that this other person killed Andy’s wife, and without knowing anything else specific to the case, we should use our general knowledge that a majority of murders are committed by 1 person and not by a group of people. That should allow us to make a probabilistic conclusion about the estimate of Andy’s guilt, and we should lower our estimation of the likelihood of the truth of Andy’s guilt.

    PS: Further, in the specific (hearsay) confession provided, the confession specified that the unnamed person worked alone (“they blamed it all on the banker husband”), which also counts as (bad, hearsay) evidence that exactly 1 person killed her. Again, this shouldn’t alone convince you that Andy is innocent, but it should adjust your estimation that Andy is guilty by at least a little bit.

  353. corwyn says

    @EL:
    1) Get over yourself. It is possible for someone to correct your mistakes without having a grudge against you. I have disclaimed this once already.

    2) “I never advocated nor defended having 0% confidence in a proposition.” You quote Holmes’ great blunder, talking about impossible things, and claim it is bayesian. What are impossible things other than things with 0 probability?

    3) Despite the fact that I included a blockquote to what I was referencing, you are going off on a rant about something completely different.

    This is your last chance to be civil with me.

  354. corwyn says

    @ 370 frank:

    Sorry Frank but it just can’t be read that way. If you replace impossible with ‘statistically insignificant’, that just means very improbable. Holmes statement becomes:

    “if you eliminate the very improbable, whatever remains, however improbable, must be the truth.”
    Surely everyone can see that this is just nonsense, right? What if what remains is very very improbable?

    All we are *ever* able to compare is various levels of probability, nothing (real) can ever be given a probability of 0 (or 1). In which case Holmes should properly be saying, “we should choose the hypothesis with the greatest probability.” Duh.

  355. corwyn says

    @376 Corwyn:

    I realized I was perhaps a bit harsh with Sherlock, a rephrasing that gets the entire idea across would be:
    “we should choose the hypothesis with the greatest probability, regardless of its prior probability.”

  356. John Iacoletti says

    Use u and /u inside less-than, greater-than symbols for underlining.

    Ok, even if we go with your gambler analogy and say that the gumball proposition is 50-50, the point is the same. In the absence of valid evidence, the claim, “the number of gumballs is even” is 50% and the claim “the number of gumballs is odd” is 50%, and the fact that I reject the first claim because of lack of evidence in no way alters the odds of the second claim.

    Now someone may come along and say that he believes the number of gumballs is odd because he read it in a book, or he thinks that we wouldn’t have trees if the number of gumballs isn’t odd, but that’s not valid evidence. It’s then a fallacy for him to say, “you rejected my claim, so what is your evidence that the number of gumballs is even?” It’s still 50-50. I still don’t believe either claim.

    I can talk about the same thing for a jury trial. In my hypothetical state of pure ignorance, knowing nothing about jury trials, starting with 50 50 odds is prudent.

    No, without evidence you should start with 1 in infinity, i.e. zero (the null proposition). It’s not even 1 in 7 billion, because you don’t know that she was actually killed or actually killed by a human being.

  357. frankgturner says

    @ corwyn #377
    I might have gone with something more to the effect of, ” when you have eliminated that which has such low probability as to be considered impossible for all intents and purposes, what ever remains is much more likely to be the truth.”
    .
    That is a bit wordy though and your gets the point across better. I think Sir Arthur Conan Doyle likely intended “impossible” to basically be “to low of probability to be considered significant” depending on his background. I believe that he did have a good background in the hard sciences.

  358. frankgturner says

    @ EL
    What you have asked is a harder question to answer than you realize. It will take me some time to come up with a full answer.
    .
    There is an old saying that you can keep in mind though, “Every profession has its own language.”

  359. EnlightenmentLiberal says

    @corywn
    Ok corywn. Everyone immediately understood what I meant, and there was no actual confusion. It would have been fine if you nitpicked my position, but you actually accused me of not knowing what I’m talking about and teaching badly, neither of which IMHO is true. I’m not interested in responding to your personal vendetta against me.

    @John
    About underlying. Weird, I thought I tried that long ago here, and it didn’t work. testing

    I can talk about the same thing for a jury trial. In my hypothetical state of pure ignorance, knowing nothing about jury trials, starting with 50 50 odds is prudent.

    No, without evidence you should start with 1 in infinity, i.e. zero (the null proposition). It’s not even 1 in 7 billion, because you don’t know that she was actually killed or actually killed by a human being.

    You make one good point. I do not recall if I mentioned “stabbed by a knife 47 times, and we have compelling evidence of that” in my previous post on the matter. You are right that’s improper to assume a human is responsible for the death without some evidence that the human is responsible for thedeath.

    However, you also make one bad point. The point about the null hypothesis is the point of disagreement. The null hypothesis used in the way that you are using it is not correct reasoning. That’s exactly what I’m asking you to give up. That’s exactly the bad kind of reasoning I’m targeting here. I don’t think I can make you give it up in this forum, but I hope to at least make you think about it.

    Again, I think the best way to think about it is as a hypothetical cliche self-interested rational agent, e.g. rational gambler. For any well-formed proposition, a rational gambler is never going to start with a literal estimation of “1 in infinity”. That’s equivalent to a literal absolute 0% estimate.

    For starters, this would cause the gambler to make bad bets. Remember, the entire point of this exercise is to assign the best estimate that one can to (epistemically) possible scenarios, in order to use those estimates to do cost-benefit analysis. Having a literal absolute 0% estimate will lead to very very silly cost-benefit analysis, and it will lead to irrational behavior.

    Also, the simple math of Bayes equation is that no amount of finite evidence could ever change a prior estimate of 0%. A 0% prior estimate is the ssame thing as “I am dogmatically opposed to the idea, and no amount of evidence can ever change my mind”. Thus, in order to have a prior estimate of 0%, you have to admit that you are dogmatically opposed to the idea and will ignore any evidence to the contrary, or you have to reject the applicability of Bayes equation and reject all of basic mathematical statistics as well.

    John, you’re still stuck in this wrong mindset of confidence rather than estimates. You need to let go of the need for absolute confidence, and you need to start thinking like a gambler. If you learn that someone is dead, it’s ok to use your prior knowledge about what percentage of death happens by murder by another human (or humans); it’s ok to use your prior knowledge about the statistical distribution of the number of people responsible for any individual murder. As a rational gambler, you should use all of your available information in order to make the best estimate that you can. That estimate is never going to be 0% aka “1 in infinity”.

    It’s irrational to ever have a “1 in infinity” estimate of any epistemically plausible scenario.

  360. Narf says

    And to get even more meta, John, if you want to show someone which HTML tags to use, just use the &lt; and &gt; HTML codes for the brackets of the tag, and the page won’t render them as HTML tags.  And to get even more meta than that, if you want to show someone which HTML codes to use, replace the ampersand in the code with an &amp; code.
    <u>underline</u> = underline

    Although the underline tag doesn’t seem to show up in the preview, in Chrome, though. Weird. Anyway, I usually prefer using the <i>italics </i>and <b>bold </b>tags over <u>underlining</u>, myself.

  361. Narf says

    Hmm, it isn’t accepting my underline tags, and admittedly it isn’t on the list of accepted tags, below this comment-entry box. How did you get it to accept them?

  362. John Iacoletti says

    <u> and </u> work for me. I didn’t do anything to explicitly enable that.

    I think the gambler analogy falls apart for a proposition that actually has no evidence. You’re assuming (rightly so) that if someone is on trial there is some evidence, but would you place a bet for any finite payoff on a number for a roulette wheel that could come up with any unbounded positive integer? What if some unidentified person came and arrested you and the only thing he said was “Alice is dead, we believe you killed her”. Without any evidence, would you assess the probability of that being correct to be greater than zero?

  363. Narf says

    You have admin rights, don’t you? I imagine it will accept whatever HTML tags you throw in.

  364. EnlightenmentLiberal says

    @John

    but would you place a bet for any finite payoff on a number for a roulette wheel that could come up with any unbounded positive integer?

    The number of plausible scenarios is not unbounded in that sense. I’m pretty sure the rules of physics are accurate, and that alone places some constraints on the size of the number of plausible scenarios. A very, very large number, but a finite number.

    This is one area of my ignorance, but I do not know offhand how one might talk about the statistical distribution “choose a random Natural Number”. However, we can talk meaningfully about the statistical distribution “choose a random Real from 1 to a googol”. The Reals from 1 to a googol are infinitely divisible, and it is a larger set than the Natural Numbers (in a formal mathematical sense).

    What if some unidentified person came and arrested you and the only thing he said was “Alice is dead, we believe you killed her”. Without any evidence, would you assess the probability of that being correct to be greater than zero?

    Of course yes. You wouldn’t !?

  365. John Iacoletti says

    I’m not convinced that the number of possible explanations for “Alice’s death” is finite. But even if it is, without any further information there is no way to even quantify what the probability might be. We don’t even know if there really is an Alice.

  366. EnlightenmentLiberal says

    @John
    I think your example is wonderful. Can we focus on that for a little bit please?

    What if some unidentified person came and arrested you and the only thing he said was “Alice is dead, we believe you killed her”. Without any evidence, would you assess the probability of that being correct to be greater than zero?

    Completely out of the blue (from your perspective and knowledge), let’s suppose a police officer comes up to arrest you for the murder of John Smith. At that moment, would you actually assert that there’s no chance at all whatsoever, a 0% chance, that there is a John Smith, who died, and some previous actions of yours directly led to his death in a criminally culpable way? I answered the question. I’m curious of your answer.

  367. frankgturner says

    @ EL
    I am sorry but I can’t answer your question. Well I can but I don’t think it would be productive.
    .
    You’ve got John’s attention, maybe that will do something good.

  368. EnlightenmentLiberal says

    @Frank
    How about a couple simple yes / no questions.

    In the post above, do you consider it likely that Matt used the words “consider”, “assess”, and “examine” with intended meanings so that they are largely interchangeable? Judging from context, sentence structure, etc., I think that likely.

    In that particular post and context, do you consider it plausible that Matt used the words “consider”, “assess”, and “examine”, with intended esoteric legal meanings which are inconsistent with the normal usage of the words – all in the same post? I consider that scenario to be highly unlikely.

    I believe that this is the scenario which are you proposing. Did I get your beliefs about the intended meaning wrong?

  369. EnlightenmentLiberal says

    @frank
    If I say “The claim cannot be examined. The claim cannot be assessed. The claim cannot be considered.”, is there a technical esoteric legal meaning for all of that together which 1- has the same meaning, and 2- has a different meaning that the everyday common-person meaning? Just a yes or no will suffice for me, but please feel free to expand.

  370. frankgturner says

    @ EL
    That depends on what you mean by esoteric. There is a dictionary definition that indicates ideas being understood by a small number of people. However, language is dynamic and fluid. Over time word meanings that are specific to an area of specialized knowledge can make their way into common speech. Many For example, many of us understood the legal idea behind “innocent” and “guilty” even though you had not been exposed to it. Sometimes it is up to the listener to recognize that they have not been exposed to a principle.
    .
    Also “cannot be examined” might apply to a specific time. So “cannot be examined” usually references not being able to examine at this current time. It does not necessarily mean “can NEVER be examined at ANY time.” That may also apply to a specific person being spoken of, e.g.: while this can be examined by others at another time it cannot be examined by you at this current time.
    .
    That is a bit of where your Bayesian reasoning comes into play. While there might be evidence that something cannot be examined by you at this current time and hence you conduct an analysis of statistical likelihood based on what you have, other evidence may later be discovered that alters that principle.
    .
    For example, earlier when discussing the idea with response to “when you have eliminated all else as impossible, whatever remains, however unlikely, must be the truth” and corwyn corrected you, you said that “Everyone understood immediately what I meant and their was no actual confusion,” my thoughts were, what evidence do you have that absolutely everyone does understand?
    .
    Based on the evidence from this board (including my own response) I would have said, “my general feeling is that the majority of people on here understood the general principle without needing correction.” It sounds like couching your terms but it demonstrates an openness to alternate evidence. If people did not comprehend and wanted to challenge it based on accuracy and content I personally would not have minded nor viewed it as a personal attack.
    .
    Something similar happened on another one of these boards when I said, “do into others as you would have them do unto you,” and Narf here stated that more accurately. If I could remember how he put it I would state that.
    .
    So if someone stated that, “It cannot be examined,” I would look for something in that you might have heard of in English class called “context clues” to indicate whether they just meant by a specific person at a specific time. If I could not determine those things I might take it upon myself to ask. I don’t consider that particularly esoteric as it can occur (and might occur) to others to ask as well.
    .
    That is sort of like “cross examining” the statement “it cannot be examined.” That can and does happen in a courtroom, that is part of why things like “cross examination occur.

  371. John Iacoletti says

    Completely out of the blue (from your perspective and knowledge), let’s suppose a police officer comes up to arrest you for the murder of John Smith. At that moment, would you actually assert that there’s no chance at all whatsoever, a 0% chance, that there is a John Smith, who died, and some previous actions of yours directly led to his death in a criminally culpable way? I answered the question. I’m curious of your answer.

    Probably not an equivalent scenario, because we’re talking about what a jury would decide. I have information about my own actions that a jury wouldn’t have. But as I said, even if the answer isn’t zero I would claim that without any evidence it’s indeterminate. So the claim should be rejected on that basis. No need to even begin trying to assess alternatives because there’s nothing to assess.