Do atheists pretend to know things they don’t know?

November 29th, 2014 in clues. Tags: , , , ,


Last post I offered some thoughts on faith, based on Peter Boghossian’s idea that faith is “pretending to know things you don’t know”. I suggested that he was mistaken in what many christians think about faith.

So, do atheists think very differently to christians on this?

Knowledge and belief

Everyone has beliefs, for beliefs are simply the things we think – from the simple and certain (e.g. “I like the colour yellow” or “I have a head”) to the deep and complex (e.g. “Jealousy is wrong” or “There is a God”).

‘Knowledge’ is when a belief is true and we have good reason to think it is true. We can know things by observation, reasoning, experience, or on the authority of someone else who knows.

The big problem when applying these definitions to a question like the existence of God, is that we may all believe we have good reasons for our beliefs, but we may disagree about what is true. Thus we disagree about who “knows”.

Science, proof, probability and belief

Science is a structured way of knowing using observation, reasoning and authority. However it doesn’t provide “proof” of very much (only mathematics and formal logic provide genuine proof), but rather “knowledge” with a given degree of probability, generally very high.

Many other things we know have similar high probability (e.g. I believe the external world exists but I cannot prove it), while many other things we know or believe are less certain. For example, we trust close friends, choose to vote for a candidate, choose careers, decide on what we believe is right or wrong, have favourite authors or musicians with varying degrees of certainty.

Our belief, disbelief or lack of belief in God are similarly uncertain, even though we can give reasons for our beliefs.

Atheism as a belief

Atheism is a belief in the philosophical sense – atheists think that God doesn’t exist or at least they have no belief that he does.

Atheism as lack of belief

Many atheists emphasise that they don’t actively disbelieve in God, they just lack any belief that God exists. But what does this mean?

It may mean that they have no belief either way about God, what many would call agnosticism. But do they therefore think that the existence of God is a 50/50 question – they lack both belief and disbelief?

I don’t think I have ever met a self identified atheist who would think that. I think they would all say they think the existence of God is not very likely. In other words, they may lack a belief in God but they certainly don’t lack unbelief!

I think therefore that most atheists actively disbelieve in the proposition “God exists” and believe the proposition “God doesn’t exist” is most likely.

Atheism as belief that God doesn’t exist

If atheists agree with the proposition that God doesn’t exist (probably), then one can ask on what evidence that belief is based. I imagine most would give arguments based on the evil in the world or the hiddenness of God – arguments which perform much the same function, except in reverse, as arguments like the cosmological or the design arguments which are used to support God belief.

Would it be fair then to suggest that this was holding a belief that they cannot know?

I don’t think so. They think the proposition “No God exists” is probably true and they can offer reasons in support. The situation is analogous to the way I hold my belief in God.

But on Peter Boghossian’s terms, they are as much guilty of this as he argues christians are.

Other common atheist beliefs

Remember I’m still using “belief” in the philosophical sense of something we think to be true. And there are many other beliefs that seem to be held by many atheists. I’ll pick just one (I have links to others below).

Atheists and the odium of religion

Religion poisons everything, Christopher Hitchens famously wrote, while Jerry Coyne is quoted as wanting to weaken religion’s “odious grasp on the world”.

Others go further. Following Richard Dawkins, some atheists accuse christians of having a delusion, Peter Boghossian says “When I speak to speak to somebody of faith, I view them as a person who really is mentally ill.” and John Loftus has said “religious faith is a mind virus” and “a public health crisis”.

Very little evidence is offered for all these claims. It seems that it’s just obvious.

But the scientific study of religion is a growing field, and has gathered significant evidence. The following conclusions are established by many scientific studies.

Christians and mental health

The evidence I can find is not all one way, but overall it shows that christian faith (and some other religious belief) is actually associated with higher levels of physical and mental wellbeing. (There is a link at the bottom to a list of studies.)

For example: In How God Changes Your Brain: Breakthrough Findings from a Leading Neuroscientist Andy Newberg says: “hundreds of medical, neurological, psychological and sociological studies on religion [show] …. even minimal religious participation is correlated with enhancing longevity and personal health”

Christians and prosociality

Christians generally contribute more to society than average (this is known as “prosociality”), as I have shown before in Does religious belief make you more moral: a case study in misusing data?.

The Science on religion blog (Is religion good or bad for the world?) quotes a recent meta study: “Contrary to the forceful assertions of some prominent atheist authors (e.g., Dawkins 2006; Dennett 2006), however, the data consistently point to a negative association between religiosity and criminal behavior and a positive association between religiosity and prosocial behavior.”

Counter examples

There are counter examples less positive towards religion. Much evil has been done in the name of religion, including christianity. Some christians behave badly. Religious belief can be associated with bad behaviour and emotional states.

But many atheists don’t discriminate. They can generalise from the bad examples that come to their minds and ignore the larger number of good examples.

Researchers discriminate between intrinsic religiosity (where people have a personal commitment to their religion) and extrinsic religiosity (where people follow a religion for what they can get out of it), and and this sometimes explains the good and the bad (e.g. extrinsic religiosity tends to lead to prejudice against strangers but intrinsic religiosity leads to welcoming strangers).

If we are going to follow the evidence, it isn’t enough to quote a few studies that seem to prove our point and ignore the many that show the opposite. And the evidence overall (as far as I have been able to find) supports the view I have presented here, and not the negative view presented by many atheists.

Defining and treating a delusion

Delusions and their treatment have a number of characteristics that don’t and shouldn’t apply to christian belief, for example:

  • “People with delusional disorder experience non-bizarre delusions, which involve situations that could occur in real life, such as being followed, poisoned, deceived, conspired against, or loved from a distance. “ (WebMD). That doesn’t describe christian belief at all.
  • “delusions commonly represent an underlying organic illness that warrants specific treatment.” (Psychology Today). What underlying organic illness are christians supposed to have?
  • If christianity was truly a delusion, then the “treatment” suggested (ridicule) is completely wrong – ” Avoiding direct confrontation of the delusional symptoms enhances the possibility of treatment compliance and response.” (Psychology Today).

Atheists and this evidence

Not all atheists agree with the “delusional” accusation, but those who do are clearly believing something they don’t know is true (because it isn’t true!):

  • Christians generally have higher mental wellbeing. Their faith is not generally a delusion.
  • Christians are more prosocial.
  • The ridicule that some atheists prescribe to combat religious belief is the opposite of what psychiatrists recommend if it was truly a delusion.


  1. I don’t think that atheism is pretending to know things that atheists don’t know, but it is as likely to be true of atheists as of christians.
  2. The commonly held atheist belief that religion is a harmful delusion, appears to be contrary to evidence, and is thus pretending to know things that they don’t know.


Please note, I have referred all through this article to those atheists who hold certain views. I know many atheists who don’t hold these views and in fact disagree with them. I wouldn’t want anyone to think otherwise.

Read more

Photo Credit: benjaminasmith via Compfight cc

Note: I do not know the person in this photo – it is just a freely available anonymous photo – and I have no reason to suppose she holds to any of the views discussed in this post.


  1. From my perspective ( which I may not explain very concisely) my atheist thoughts went along the lines of ‘people have constructed a purely fabricated idea and I will have no truck with it’. I was always exaspected by the argument that the position I was taking was one of faith or belief in itself. Someone had invented an imaginary idea and I thought that they were delusional.
    If I came to visit you in Sydney and said ‘look, I’ve brought my gold and sparky unicorn with me, here she is’, You would say ‘Eva, there is no unicorn there but would you like a coffee and a scone?’ And we would have a nice afternoon. Your denial of my pet would not constitute a whole belief system on your part- it would just be a simple ‘no it’s not’.

    My atheism did then morph into a ‘religion is bunk and believers are delusional and possibly ill’ which I am ashamed of now of course.

    Basically, I don’t think denial of something that you believe to be invented constitutes any kind of ‘belief’. Once you design a personal philosophy around it though, it moves into other areas.

  2. For me you nailed it exactly in your conclusions. The problem isn’t atheism nor religion. The problem is certain atheists and religious believers who get a tad (or more!) too enthusiastic about their worldview.

  3. I think Dawkins only used ‘delusion’ as a shorthand and coined the new word ‘relusion’ for what he meant. There’s still some genuine nastiness about that, but he meant something different.

    When I speak to speak to somebody of faith, I view them as a person who really is mentally ill.

    If Boghassian said that (I have not read his book, but I trust your portrayal), he earns a loss in respect for pathologising people with a different opinion and for using ‘mental illness’ as a playground insult.

    Eva, I’d actually consider my denial of the existence of fairies, unicorns and the typical atheist facetious placeholders for God as a belief on my part, one that I hold rather strongly. But I wouldn’t call it a belief system. Still, my disbelief in cryptids and the paranormal are less important to me than my belief in God, yet it means quite something to me.

    Of course, this comes down to the not-so-important semantics of ‘belief’ and ‘believe’.

  4. Hi everyone, thanks for comments. As stated in the post, I have used “belief” in the philosophical sense of something we think is true, and it doesn’t have “faith” connotations necessarily. I have written it all to parallel Boghossian’s views on how christians think with how some atheists think.

  5. This is one of those rare occasions that I happen to agree with you, unkleE. 🙂

    I would certainly agree that Christians are, ” prosocial ” Having attended various churches for over 50 years, I found many attendees looked at their church as their social center.

    “The evidence I can find is not all one way, but overall it shows that christian faith (and some other religious belief) is actually associated with higher levels of physical and mental well being”

    I would be curious to know if there have been studies about other social groups like social clubs / organizations. Are members of non-religious social organizations also associated with higher levels of physical and mental well being. Although I have no evidence for this, my initial inclination would be, “yes” . Your thoughts ?

  6. Hi Ken, I think you are right about other social groups, though I don’t have a reference I can put my hands on. Three of the most important contributors to wellbeing are a close circle of friends, meaningful (especially volunteer) work and serving a cause greater than oneself. Now religious attendance and activity can contribute in all these areas but I can imagine (and think I have read somewhere) that other social groupings can contribute to some of these, though perhaps not as much.

  7. If you don’t mind me focusing on a small part, I’d like to throw in some extra details about scales that operationalise styles of religiosity. This animal has some of that, plus an opinion on them. 🙂

  8. All right. You know about the religious orientation scale of Allport and Ross and the Quest religiosity scale of Batson.

    Allport and Ross valued intrinsic religion (I), ‘sincere, personal belief’, over extrinsic religion (E), ‘a utilitarian approach to religion’. Batson on the other hand valued Quest spirituality (Q) over the other varieties. It is useful to keep these biases in mind with those scales.

    Also, these scales have their origin in a Protestant Christian context. I’m not sure any more how much it held for other religions, unfortunately, but this background will have an effect.

    The relation between religious orientation and prejudice is a little more complex than just a high score on Quest or intrinsic religion versus scoring high on extrinsic religion. It has been found that fundamentalist Christians tend to have an “all of the above” approach to intrinsic and extrinsic religion (but not Quest religion). This is called indiscriminately pro (IE) and correlates more strongly with social prejudices than extrinsic religion. The resulting relation with prejudice from stronger to weaker is something like:


    But I think this is only part of the story. I think this generally holds for societal bias, but it has also been documented (by historians) that New Age people (who can be expected to score high on the Q-scale) are prejudiced towards. So there seems to be some intergroup prejudice at play here. It would be interesting to see if that can be operationalised.

  9. The problem with the main body of your argument, which essentially concludes that atheists think God *probably* isn’t real, and that this is equivalent to theists who say that God *probably* is real (let’s ignore those who are certain on either side, because I think we can agree that both of those groups are unjustified in their certainty) is that it relies on a shifting of the burden of proof.

    For all claims, the burden of proof lies with the person making the claim. If I claim “God does not exist”, I need to prove that – and, as you noted, I cannot do so. *But*, atheists generally are not making that claim – or any claim, in fact. Atheists are *responding* to the claim that “God exists”. We take a look at the best available evidence and conclude that the claimant’s burden of proof has not been met. We then reject the claim “God exists” as unproven, expressing a belief that “based on the evidence available, it is more likely that God does not exist than that He does”. That is very, *very* different to simply claiming that “God (probably) does not exist”.

    Atheists do not need positive evidence that God does not exist in order to reasonably reject claims that don’t meet their burden of proof. The position that “I do not believe you” is the *default* position for any claim that is not supported by sufficient evidence, and the question of God’s existence doesn’t deserve special treatment in that regard. It is only once enough evidence to support the claim’s truth has been presented that “I do not believe you” becomes unreasonable denialism, rather than rational skepticism.

  10. Hi Cheomit, thanks for that comment. I would appreciate the opportunity to explore it further if you are interested. I’ll make just two points for starters.

    1. I was using the term “atheist” in the sense used mostly by philosophers – e.g. the Stanford Encyclopedia of Philosophy, which I think is considered to be a reasonable authority, says: “‘Atheism’ means the negation of theism, the denial of the existence of God.” I know some atheists like to define their view as simply lack of belief, but I think it is best to stick with the one given by Stanford.

    2. In the post I suggested that atheists who say, as you indicate here, that they have no belief that God exists, probably do have a belief that God doesn’t exist. Certainly most atheists I have “met” on the internet seem to quite strongly believe and advocate for the proposition “God probably doesn’t exist”.

    I wonder if I can test that on you please. Would you say you give equal probability to the two propositions “God exists” and “God doesn’t exist”? Or a different probability?

    Thanks again for raising an interesting question.

  11. Hi unkleE,

    Sorry for the slow response – holidays and whatnot! Hopefully you’re still interested in a discussion.

    Let me address your question first. It’s not as simple as it might seem, though, because it depends on how you define “god”. If we’re talking about a purely deistic god, one who created the universe through some “big bang” process, and who butted out thereafter, then I’d say the probability is approximately equal, simply because we have (and likely can have) no evidence to sway us in either direction. There are some valid concerns which might shift it slightly from equal (where did god come from, for example), but I think that neutrality is a reasonable position. Of course, I’m talking *epistemological neutrality* – that is, neutrality with respect to claiming knowledge. By definition of having no evidence either way, we should not believe either “a deistic god exists” nor “a deistic god does not exist”.

    If, however, you want to talk about an intervening god, like the Christian God, then I would say that the probability of his existence is lower than the probability of his non-existence. First, there are empirical claims made about that type of god – its nature, what things it has done, reported miracles, and so on – which have not met the burden of proof to which I originally referred (or which have been actively disproved), and should therefore not be rationally believed. There are also valid points which are directly against the existence of such a god. None of these things add up to *disproof*, but they do constitute reasons to justify a shift in probability.

    It’s just like any other claim. If there is no evidence to support the claim being made, you reject it until such evidence is forthcoming. If nobody bothers to look for evidence, and there’s no reason to justify a belief that it’s false, then that’s likely as far as you go. If, however, billions of people have been searching for that evidence for a couple millennia, and have come up empty handed, then the probability that the claim is true goes way down. If there are facts or arguments which contradict the claim, they too lower the probability of it being true. As you said, outside math and logic, we never get to zero, but we can definitely move that way – and atheists, I think, are entirely justified in doing so.

    Let me now turn to your responses. Firstly, I’d say that the SEP page you linked paints a rather more complex picture than just “‘Atheism’ means the negation of theism, the denial of the existence of God.” Other dictionaries (such as the Oxford) also offer the “lack of belief” definition; and even the definition of “belief” varies, including active disbelief alongside withholding judgement. Furthermore, in a philosophical sense, the opposite of believing “X” is believing “not X”; disbelief, on the other hand, is believing neither “X” nor “not X”.

    Regardless of that, though, I think it’s important to interact with how people *use* the word, rather than how it is defined. Dictionaries and academics take their time catching up on changes in use, after all. Most of the people I listen to and talk with distinguish between “hard” and “soft” atheism, or “anti-theism” and “atheism” (the formers being “there is no god”, the latters being simple lack of belief). To be clear, they don’t always make the distinction *explicitly*, or using those terms, but if you probe their thoughts, you usually find a split along those lines. I’d also say that faced with claims farther along the deistic side of the scale, most tend strongly toward the “soft” position; once you start getting toward a more concrete, interventionist god, though, the positions sometimes harden – which, as I said, I think is justifiable (although truly hard atheism – “there is no god”, even if talking about a *specific* god – is, I think, generally not).

  12. Hi, thanks for that response, there are some good things to chew on there.

    1. I have no difficulties with your position regarding a deistic god. I think the evidence is quite good for such a being, but I think your ” we should not believe either “a deistic god exists” nor “a deistic god does not exist”” is fair.

    2. While I don’t agree with your view re the christian God, you have made yourself quite clear about what you have concluded. But I think “there are empirical claims made about that type of god …. which have not met the burden of proof …. There are also valid points which are directly against the existence of such a god” is a proposition with quite definite content, and as such requires evidential support – which you have indicated you belief you can provide. So aren’t we equal here, except perhaps my belief in God may be stronger than your disbelief?

    I would be interested to see a quick list of the evidence and arguments you believe either are falsely claimed to point to God, or point away from him.

    3. But I find myself much more in disagreement, when you say: “billions of people have been searching for that evidence for a couple millennia, and have come up empty handed”. On what criteria, what evidence do you make this statement? Billions of people don’t believe they have come up empty handed, and many (hopefully I am one of them) believe we can state that evidence clearly. Wouldn’t it be truer to say that though these billions believe they have evidence, you don’t accept that evidence as being true and/or relevant? Which of course is a very different statement.

    4. Finally, I agree with you that we should go with the way people use words. But in this case, there is no consensus. Most theists use “atheist” in the “old” way, many atheists use it in the “new” way. Now normally I am happy to let people self identify, but I think there is a little trick going on here sometimes. I have seen atheists argue vehemently against God’s existence, then when asked for the evidence for the very definite propositions, they say they don’t need evidence because they don’t have any burden of proof because they are not expressing any belief – and then go on making negative propositions about God. If atheists wouldn’t play that game (and of course not all do) then I would be happy to use the word in the way they prefer, but as it is I think it actually confuses.

    Well that’s enough from me. What do you think about all that? Thanks.

  13. Hi unkleE,

    Sorry again for the slowness of my response. I’m back around reliable internet again, so hopefully we can keep this going a little better now!

    1) I’d love to hear your evidence for a deistic god. Almost by definition, we cannot know if such a thing exists – because it would exist outside of any realm that we have access to (i.e. the observable universe) and it would have no impact upon our universe beyond its initial creation. I find it difficult to conceive of any evidence to support a conclusion that such a being exists (or does not exist, for that matter).

    2) Sure, I think can defend those positions. Just at the simplest (and thus most contentious) level, and looking at Christianity: there are no verified miracles, the scriptures are full of internal contradictions, and the existence of Jesus the *man* (let alone Jesus the divine being) is questionable. As for counterevidence, there’s the problem of evil, the contradiction between scripture and scientific evidence, and the lack of evidence for divine intervention where we would most expect to see it. None of these problems (or any others) *prove* anything – and some of them are easily swept away depending on what, precisely, one believes (e.g. you can solve the problem of biblical historicity if you believe it’s just allegorical) – but it’s not unreasonable to conclude, in light of them, that the probability is not 50:50.

    But all this, I think, distracts from what I think is the key point here. If the question is “does God exist?”, you would answer, I think, “yes”, or “very probably”. I, on the other hand, would answer “maybe, maybe not”; I would *not* answer “no”. The strength of my “disbelief” is zero, because I *do not* have any belief that god does *not* exist. The question “what is the probability that god exists” is an entirely separate one, and I can answer “I think it’s more likely that he does not exist” without contradiction because the question asks me to consider the evidence of God, not to consider God himself.

    3) Billions of people have come up empty handed insofar as finding evidence that is sufficient to justify the rather extraordinary claim “God exists”. There are mountains of “evidence” that wouldn’t stand up to the scrutiny of a traffic court, but having a mountain of poor-quality evidence (like personal experience or revelation, flawed arguments, unsubstantiated miracles, etc.) don’t add up to a convincing position. The very lowest bar, I think, is Loftus’ “outsider test of faith”, which asks you to consider the evidence or argument as if it were being made by someone of another religion. Only if it would *still* be convincing to you should you consider it evidence in support of your own position – and I’ve not come across anything that meets even that low bar.

    When we refer to evidence in *any other situation*, I think it’s fair to say that most people mean something like “evidence which is sufficient to justify, to a reasonable person, the claims being made”. There’s lots of “evidence” that suggests homeopathy is an effective medical intervention, but since none of it rises to a reasonable level, it’s entirely fair to say that there’s “no evidence” which justifies the use of homeopathy. In the same way, there is no evidence which justifies belief in God – because none of it meets even a reasonable standard, let alone the extraordinary standard required to justify such an enormous claim as the existence of God!

    4) Let me ask you about these atheists arguing strongly against the existence of God. First, let’s put aside the titles of debates – they’re usually framed as “does God exist?”, and by nature take on a “yes/no” format, which doesn’t lend itself to “we don’t know, but there’s no good reason to believe that he does”. It skews the position and depicts the atheists as active disbelievers from the beginning.

    But when you say people argue against the existence of God, do you mean they put forward arguments which are intended to prove that God doesn’t exist? Or do they, rather, argue *against* the arguments made by the believers? I think the most common “anti-god” argument might be the problem of evil – but that does not at all disprove the existence of a god, it simply highlights an apparent contradiction with the notion of an omnipotent and benevolent one. More commonly, though, I see people arguing against terrible arguments – pointing out the flaws in the cosmological or teleological arguments, trying to explain why personal experience counts for next to nothing, or explaining why this or that evidence for some miracle doesn’t meet its burden of proof.

    Arguing against bad arguments, or pointing out flaws and contradictions in someones ideas, doesn’t mean you’re arguing that their conclusion is wrong, though – it’s arguing that their conclusion is *unjustified*. And that’s the whole point – it’s the difference between “you’re wrong, god does not exist” vs. “I don’t know, and neither do you – but you shouldn’t believe if those reasons are the best you have”. It is, I think, a confusion between what we claim to know, and about what we should believe.

  14. Hi Cheomit, no worries about slowness, I think it is good to be leisurely, especially at this time of year.

    1. I think the evidence for a deistic God is very clear, but I can only give a very brief outline here. To me, most questions in life are matters of probabilities. We select the option that is more probable than any other option, based on the evidence. If a hypothesis doesn’t explain the evidence, then it has low probability.

    The origin or cause of the universe, as expressed in the standard cosmological argument is one example. When I have discussed this with atheists, they mostly say they don’t know how the universe got here, which makes atheism a poor hypothesis in this case. The fine-tuning argument can be expressed formally in even clearer terms that show that the only possibilities are design or no design, with the latter either chance or determined. Chance is an impossibly improbable option, and cosmologists mostly believe determined isn’t the case. So strong are these conclusions that the only option seriously considered is the multiverse, which then requires an explanation, and we get the same problem. Design is clearly the least improbable option. Other arguments come from human reasoning and ethics and lead to similar conclusions. If I put all those facts into a bayesian probability analysis with some notional probabilities, God is by far the most probable option. It seems very clear to me.

    2. It depends what you mean by “verified miracles” – how would you regard a miracle as “verified”? My response is to again use probability (Bayes again). Surveys suggest that about 300 million people believe they have experienced or observed a healing miracle. Even if an enormous percentage are urban myths or have insufficient evidence, that’s still a huge number of miracle reports to refute or verify. People have attempted to assess a small number of these. (e.g. The Lourdes medical commission assessed a few thousand claims that had some evidence and found that about 70 were sufficiently verified to be considered as miracles. A medical researcher has documented another 10 well-documented miracles that cannot be explained medically. A university researcher found statistically significant improvements after patients were prayed for.) The rates of recovery seem to be way beyond spontaneous remission, and a Bayesian analysis suggests that these reports dramatically increase the probability that God exists. See Healing miracles and God.

    The internal contradictions are only an issue if an argument is presented that depends on the scriptures being totally reliable, but I don’t do that, I only require the NT to be what the expert historians say it is. And the same expert historians say there is no doubt that Jesus lived a life more or less as outlined in the gospels (though of course not including the supernatural – they generally put the supernatural to one side). So this evidence is only questionable if you are unwilling to accept what the experts say.

    However I agree with you that the problem of evil is a significant reason to disbelieve in God. But when I consider all the evidence, I find the problem of evil is significantly outweighed by the evidence from the universe, human reason and morality, human experience of miracles, and the historians view of Jesus.

    So far from “Billions of people have come up empty handed insofar as finding evidence that is sufficient to justify the rather extraordinary claim “God exists”.”, I would argue that billions of people have very good reason to believe, and it is the minority who reject that evidence who are empty handed. I’m sure you won’t accept that, but it is a very strong case to me.

    “when you say people argue against the existence of God, do you mean they put forward arguments which are intended to prove that God doesn’t exist?”

    All I mean is that atheists make statements or propositions, and naturally those statements tend to express negative views about God. If they want anyone to take notice of those statements, they need to offer evidence for them, just as theists should about their statements. The only way I can see anyone has no “burden of proof” is if they don’t make statements, or their statements are very personal and weak in their impact (e.g. ‘I don’t like the idea of God’, or ‘God makes me happy’). So I guess my simple test would be to ask any atheist who says something about God: “Do you want me to take any notice of that statement?” If the answer is “no”, then we can ignore it and move on, but if the answer is “yes”, then I can reasonably ask “Why?” meaning what evidence do you offer?

    I think that’s enough from me. Over to you. Thanks for the opportunity to discuss.

  15. 1) “I don’t know” is a perfectly reasonable answer to *any* question. It’s also not the responsibility of atheists to propose an alternative model for the origins of the universe – that’s a job for cosmology and theoretical physics, which (like atheism) doesn’t presuppose *anything* about the existence of God (and if they find him, I’m sure they’ll let us know). Atheism does not presuppose *any* solution to the question, because it doesn’t *need* to (we can say “I don’t know how the universe came about, but let’s try and figure it out!”, and that is *perfectly consistent* with an atheistic view). A theistic worldview, on the other hand, proposes that “God did it” – but is that in fact an explanation, or is it just patching over the question in order to make people feel better? What, precisely, does “God made the universe” explain? Why did he do it at all? Why did he make it as it is? How did he do it? And, of course, how did God come into existence? If God could simply come into existence why could the universe not do so? If God is necessary for the beginning of the universe, why could not some non-intelligent precursor to a multiverse be necessary, thus negating any need for God? That’s a long list of questions – and probably only scratches the surface, since it’s off the top of my head – that are important, and that the theistic worldview does not actually answer any better than cosmology addresses the reasons for the Big Bang. All any of us can say is “the universe exists” and we do not know why. Some people *pretend* to know why, but that doesn’t make their claim to knowledge true.

    Also, since we’re talking about Bayesian reasoning, the mathematical support for the multiverse theory, which is slim support but is *at least* as good as any theistic proposition I’ve come across. But, if the math supporting multiverse theory does a substantially better job of describing and making accurate predictions about our actual universe than do any of those theistic propositions, it would be reasonable to argue that there’s better evidence for a multiverse than for intelligent design.

    As for the teleological argument you’ve presented, I’ll try to be brief. P2: Life *as we know it* may not have arisen, but that does not mean *life* or *intelligence* would not have done so. Post hoc reasoning. P3) Not designed does not automatically mean chance as you present it. The constants may only be able to take a very limited number of values for some physical reason, or they may be inextricably linked such that if you know one (and you know enough physics) the others can all be calculated. There may be some type of “natural selection” which results in “unfit” universes collapsing in on themselves. This premise demonstrates a lack of imagination. P4) You have no evidence to support this conclusion, since you have no access to anything outside our universe – therefore you cannot say that the existence of our universe is not the result of the laws applying to whatever is outside of it (e.g. multiverse theory). This is an appeal to ignorance. P5) Again, you have no evidence to support a conclusion of “not chance”, and you also appeal to the post hoc fallacy described in my response to P2. And again, multiverse theory can deal with probabilities that “ridiculous”, an appeal to which is an example of the lottery fallacy (“what are the chances that *I* won the lottery/live in *this* universe” vs “what are the chances that *someone* won the lottery/*some intelligent being* lives in *some* universe”). P6 therefore fails to follow, because your premises are not successful.

    2) Verified miracles would require solid evidence that an event actually occurred. That means you need actual evidence that a completely naive archaeologist or forensic investigator could draw a reasonable conclusion from. Most miracle claims lack anything like this – although obviously that’s not always true (you’ve linked a few documented examples), but it’s a critical starting point.

    Next, you’d need to actively rule out *all* known natural explanations – which for medical cases is extraordinarily difficult to do, given how amazingly many things we *know* that we don’t understand about the human body (let alone the long list of things that I’m sure we’re entirely ignorant of!). In medicine, you’d want something dramatic that has never been observed before and simply cannot be explained by anything we know, or are reasonably likely to discover, about the body – regrowing an amputated limb being a good example.

    And finally, you’d need to demonstrate that there’s no *unknown* physical mechanism at play, which is admittedly a nearly impossible bar to cross, but given that miracles are *supposed* to be impossible events (rather than just one-in-a-millon or so, like the Lourdes “healings”)… Alternatively, you’d need to actively demonstrate a supernatural mechanism (which science currently has no mechanism for doing). At the *very* least you’d need to demonstrate a strong, direct and exclusive link between, say, prayer and the recovery (e.g. every time you pray for someone with a particular condition in a specific way, they are healed; but they never recover otherwise – or at least the difference in rates of cure is *huge*, not just statistical noise, from poorly designed trials or the results of a single study). The last one wouldn’t *prove* prayer is working like you might think, but it would at least mean it was worth looking into, and would provide significant credibility to the claims.

    The biblical issues, of course, can be dealt with, as I noted and you’ve done; and as for the historicity of Jesus the man, it *is* an open debate among scholars. Certainly, the mythicist position is weak – and personally, since I lack the required knowledge, I’m going with the consensus view that Jesus the man *did* exist, but that doesn’t mean it’s a settled issue. Also, skipping ahead just a little, the historical view of Jesus the man has *no* bearing on the supernatural claims (i.e. the existence of Jesus the divine) – historians *cannot* verify miracles (because history must appeal to the most likely explanation of the facts available, and by definition a miracle is *always* the *least* likely explanation).

    Furthermore, when you’re attempting to balance out the problem of evil, how is that not simply admitting that the view of God as omnipotent and benevolent doesn’t work? If the problem of evil cannot be solved, but can be balanced out, doesn’t that imply a failure of the proposition? Also, the problem of evil is *expected* under a naturalistic worldview, and we can readily explain the universe, human morality, miracles and the history of Jesus without any appeal to the supernatural at all – why does that then balance out in favour of God?

    Finally, back to those billions of people – and the handful of others. I’m not even going to address the billions again – as I’ve said, their reasons would fail to convince a traffic court if the question were *anything* else (and again, I think the outsider test for faith is the lowest reasonable bar to set for a claim of this type – if you’ve got one that you think passes that test, I’d love to hear it). But those atheists being empty handed is the entire point. It goes back to my original response to your post – atheists are *responding* to the theistic claim, and therefore do not require evidence that God *doesn’t* exist, because that’s *not the claim being made*. The atheist claim is “there is insufficient evidence to justify belief in God, therefore we should not believe” – the burden of proof lies with those claiming that there *is* sufficient reason to believe.

    3) A burden of proof lies with the person making a claim. A claim could be “God exists”, or it could be “God doesn’t exist”. Atheists generally reject *both* those claims, on the basis that there is insufficient evidence to support either one. Now, you could, perhaps, argue that “there is insufficient evidence to justify belief in God” is a claim that needs to be defended – and atheists *do* defend that position. But it’s still a shifting of the burden of proof to make that claim, because the burden of proof needs to lie with the *positive* position (there *is* evidence); the negative position *there *is not* evidence) is, again, the *default* position that we should take toward *any* claim until presented with evidence by people making the claim. It’s like a courtroom, where we presume innocence (i.e. no crime took place) *until* the prosecution makes a sufficiently convincing case for guilt (the standard of evidence required depending on the nature of the crime in question). Likewise, we should *presume* that any other proposition is false *until* it is proven to be true – and that should extend to belief in the existence of God. Assuming that some claim is false until we are presented with evidence sufficient to justify acceptance of the claim is the default rational position, and the burden of proof lies with the person making the claim. Rejecting a claim does not require adopting a burden of proof, because rejecting a claim does not entail adopting the opposing claim.

    Consider an analogy (stolen from Matt Dilahunty). Let’s imagine we find a jar of jelly beans, and I say “the number of jelly beans must be even”. You might respond “how do you know?”. I might reply “I have a hunch”, to which you may respond “well, I don’t believe you”. Does that mean you think that the number of jelly beans is odd? Do you need to defend the proposition that the number is not even? Or are you adopting a position of not believing that the number is either odd or even, awaiting further evidence?

    It’s the same with every claim of the kind “X is evidence for the existence of God”. The burden of proof lies with the person making the claim, not with the person who, by default, rejects it. If you think X is evidence, you need to convince me as to *why* it is evidence; I don’t need to prove that it’s *not* evidence. If I’m a reasonable, rational person, or if I want to convince *you* as to why your claim doesn’t hold up, then I’ll give my reasons for rejecting it (too weak to support such an extraordinary claim, or easily explained by natural means, or whatever the case may be) and we can discuss those – but I’m not actually required to do even that! Certainly, it makes for a much better discussion if I explain my reasoning, but since you’re trying to convince me of something, I have no obligation to explain why I’m not convinced (unless I want to convince *you* as to why your reasons are unconvincing!). Again, at *best* the largest burden of proof that an atheist is required to adopt in a discussion is to explain why *your* reasons for belief are unconvincing. We don’t need to prove your reasons wrong, and we certainly don’t need to show that God doesn’t exist – all we need to do is show that your reasons are not good *enough* to justify belief.

    That was a long reply – and I think a lot of it is deviating from the original point about the burden of proof – but I hope you found something interesting there. I’d be more than happy if you want to pick out only the parts most relevant to that original discussion to bring things back on track a little!

  16. Hi Cheomit, I didn’t expect we’d suddenly see eye to eye, so I’ll just continue to explain how I see things.

    ””I don’t know” is a perfectly reasonable answer to *any* question.”

    Yes it is. But a hypothesis that has an answer is better than one that doesn’t (at that point at least). Of course “God did it” doesn’t explain how he did it, but at least it explains the cause. That’s the important question for the discussion here.

    ”the mathematical support for the multiverse theory, which is slim support but is *at least* as good as any theistic proposition I’ve come across.”

    The mathematical support for the multiverse isn’t the important question here. It is important for cosmologists, but we are discussing metaphysics. The important question is what these scientific ideas tell us about the question of God. And the multiverse still has to be explained – how come we have an amazing multiverse that just happens to crank out 10^500 universes that all have different properties, and which include our rare and amazing universe? The multiverse appears to be as fine-tuned as our universe that we are trying to explain.

    ”Life *as we know it* may not have arisen, but that does not mean *life* or *intelligence* would not have done so. Post hoc fallacy.”

    It is different to what you apparently think. Of the absolute zillions of possible universes, almost all would be just composed of hydrogen, or just composed of helium or just a thin soup, etc – no chemistry, no complex molecules, no habitat, nothing from which life can form. If you can imagine life under those circumstances then most physicists can’t! And anyway, it would still be the case that our universe is spectacularly unusual and suitable for us.

    ”Not designed does not automatically mean chance as you present it. The constants may only be able to take a very limited number of values for some physical reason, or they may be inextricably linked such that if you know one (and you know enough physics) the others can all be calculated.”

    Yes, that is true, that is one of the options I suggested. If not designed then either chance or the result of some natural law or process. But the physicists don’t think that is so, and anyway, we are asking why the laws are as they are – saying because of some law is just circular and no explanation. There just doesn’t seem to be any reasonable explanation except that it was designed.

    So I think all your statements about this argument haven’t really fully grasped the argument or the science. I would be happy to go through them to argue that in detail if you like.

    ”Verified miracles would require solid evidence that an event actually occurred.”

    I agree with you. All that has been done, in some cases. In the end, we can say that something very unusual happened. Whether we judge each case to be more likely a miracle or some other unusual cause is a matter for each person. But I think it would be a pity to miss out on some extremely useful information because we had set up arbitrarily high standards of proof. Proof isn’t possible, but probability is.

    ” the historicity of Jesus the man, it *is* an open debate among scholars. “

    I’m sorry, but this simply isn’t the case. It is a quite settled question among scholars. There are thousands of New Testament and classical historians, and the mythicists are limited to 2 maybe 3, and the two I know of are not working in a credible academic organisation and generally don’t have the respect, and certainly not the agreement, of their peers. It is only an open debate among non-academic non-expert atheists.

    ”the historical view of Jesus the man has *no* bearing on the supernatural claims (i.e. the existence of Jesus the divine)”

    I agree with you that we should distinguish between the two, but the history does have some bearing on the belief. If there was no historical evidence, it would be hard to have any belief. But granted there is historical evidence, we then have the task of explaining that evidence. That is a matter for each of us, but a good case, in my mind a compelling case, can be made for Jesus’ divinity.

    ”Let’s imagine we find a jar of jelly beans, and I say “the number of jelly beans must be even”. You might respond “how do you know?”. I might reply “I have a hunch”, to which you may respond “well, I don’t believe you”. Does that mean you think that the number of jelly beans is odd? Do you need to defend the proposition that the number is not even? Or are you adopting a position of not believing that the number is either odd or even, awaiting further evidence?”

    This is a great example, thanks for it. My response is simple. We need to first determine what “I don’t believe you” actually means. Is it claiming there is really an odd number? If so, you can ask why I think that. And since there is a 50/50 chance you are right, I would be foolish to make such a statement. But if the statement means simply I don’t think you can be that confident, then that is a correct statement that can be justified by mathematics.

    So as for atheism, the important thing isn’t the words but what they mean and whether they can be supported by evidence and argument.

    ”If you think X is evidence, you need to convince me as to *why* it is evidence; I don’t need to prove that it’s *not* evidence.”

    I disagree here. If I produce as evidence in court fingerprints found on the gun, and you as the defence lawyer want to reject them you really do have to explain why fingerprints on the gun don’t mean what I have argued. So if I say (as we have discussed) that the teleological argument is evidence for God, then if you want to win the debate or convince me I’m wrong, you very much have to argue why the premises are wrong – which you have attempted to do.

    I have replied at too much length too, but that seems to be the way these things go! Best wishes.

  17. Hi UnkleE,

    All fine by me! I’m up for a discussion, wherever it leads. Just didn’t want it to seem like I was trying to derail things on your page!

    Since I haven’t found the italics tag, I’m going to respond to each of your italicised headings by number.

    1) “God did it” isn’t an explanation – it’s much more like a hypothesis masquerading as an explanation. If “God did it” was an explanation, it would actually *tell* us something of value; it would contribute substantially to our understanding of the universe; it would allow us to make predictions about reality. When you get down to it, “God did it” just means “I have no idea what happened, but I know a guy who does”.

    2) If nothing else, a multiverse can be treated as equivalent to God, simply as an unintelligent universe-creator. For example, the easiest hand-wave is to say that the conditions that lead to the multiverse are a logically necessary feature of whatever higher reality exists outside our own universe – just like a theist could claim that God is a necessary being. It’s probably just as dishonest a claim to make either way (given that neither person could actually justify their claims), but it’s entirely equivalent and equally unfalsifiable. And if you stop at either point (God did it/multiverse did it), given our current knowledge, you’re using special pleading in order to avoid an infinite regress of “well, what caused *that*?”. Either way, though, the multiverse theory offers *at least* as good an explanation of why our universe exists and *appears* to be fine tuned as does “God did it”.

    Further, the math is relevant as evidence for the metaphysical proposition. If the math which leads to a multiverse does a better job of otherwise describing known reality than do theistic claims to intelligent design, then on a Bayesian view you have to conclude that a multiverse is more probable than an intelligent designer. Again, I don’t know anything like enough to be sure if that’s the case, so I won’t actually plant a flag on that claim – but it’s important to consider the possibility.

    Regarding your claims to probability, though, I need to again object. Firstly, you *cannot* claim that there are 10^500 possible universes. We’ve got a sample of *precisely one* universe to work from, and we cannot say what things could or could not be different, let alone by how much. Maybe there is *no other possible value* for the mass of an electron, given the nature of reality. It goes back to the point that you’ve no justification for concluding that our own universe is not as it is by physical necessity.

    Let’s assume, though, that there is a multiverse which, by its nature, spat out 10^500 universes (just like the claim that God, by his nature, created ours). That means that any event which occurs more often than, say, 1 in 10^498, is almost certain to occur (if i remember my stats, if you want to catch a 1/x event, you need n = 3x samples to reliably detect it in a trial). Again, we have no idea how common life even in our own universe is – let alone in any others that might exist – but that simply means that you don’t get to claim that it’s impossible just because you can come up with a number that makes it seem absurd to our tiny human brains.

    3) So, firstly, we know *for a fact* that there is at least one universe in which life *as we know it* can exist. We have no particular reason to think we’re special, though, so it’s not at all unreasonable to think that life could exist elsewhere, or in other forms, than our own. If other universes, with very different physical laws, do exist, then there’s no reason to believe that intelligent life would take on a form we would even hope to recognise as such. The common presentation of God (not one which I’ll ascribe to you, since I’ve not seen you claim it, but one which is common nonetheless) as a disembodied mind even explicitly admits that intelligent beings could exist that bear no resemblance to ourselves – which me might not even recognise as intelligent or beings if we stumbled across them, given our limitations!

    And again, if there are 10^500 universes, intelligent life and universes capable of supporting it don’t have to be even remotely common in order to exist. Further, we already know of one that does, and it’s reasonable to assume that some of those 10^500 universes would have conditions substantially similar to our own (physical constants all shifted by +/-0.00001% or some such fraction) and would likely support life like ours. Again, it’s *at least* as good an explanation as intelligent fine tuning.

    4) Firstly, physicists do not generally claim knowledge about what, if anything, goes on outside our universe, or even during the very earliest stages of the big bang. They *propose* hypotheses, some of which are a consequence of the best mathematical models we currently have, some of which are the result of messing with the math to see what happens, and some of which are little more than wild conjecture. But that’s what science does – proposes wacky ideas and tries to figure out if they’re true or not. Very different to claiming that we already have an answer.

    Secondly, physical laws are *descriptive*, not *prescriptive*. The laws we have are just representations of what we’ve found out about the universe. The universe *is as it is*, and our laws describe that universe to the best of our ability. Circular, sure – but entirely true. Asking *why* they are that way is a worthwhile question, and one to which we don’t – and may never – have an answer. Just because we can ask a question, doesn’t mean there *is* an answer, let alone one we have the ability to actually discover. And, again, simply claiming to have an answer doesn’t mean we’re correct.

    I’d be happy to discuss the teleological argument further – I don’t claim to be an expert, but am entirely willing to learn and to do my best to impart my own understanding, limited though it may be! Perhaps separately, though – I’m long winded enough without multiple arguments at one time! If you care to respond to these previous four points, that’s fine, and I’ll let you have the last word on the issue for now (unless you say something I find particularly egregious!).

    5) Often (of course not always) with miracle claims, there’s not even evidence that *something* happened, let alone something unusual. And unusual doesn’t imply miraculous, either – which I think is a big problem in how people view miracles. At the most mundane level, you have people who pray for help finding their keys, and claim that it’s a miracle when they do – which is obviously absurd. At the other end, though, you have people claiming that winning the lottery or having a disease undergo spontaneous remission is a miracle – even though we know that someone wins the lottery every week, and spontaneous remissions take place in some small number of cases, regardless of prayer or piety – or even which religion they subscribe to. A miracle can’t just be a rare event – because with 7 billion people currently on the planet and thousands of years of recorded history, pretty rare things are going to have happened a *lot*! In order to be convincing, surely a miracle needs to be an *impossible* event – something that breaks our best established scientific models of the universe, something that is indisputably supernatural in origin. Instantly regrowing a limb in response to prayer; or complete resurrection and regeneration of a week-old, half-eaten corpse – *those* types of things would constitute events which are probably impossible through any reasonably likely natural means, and might justify looking toward the supernatural. Recovering from heart surgery or finding out you’re unexpectedly cancer-free are everyday events, and even though they might *seem* miraculous if they happen to me, it doesn’t mean that anything more than the law of large numbers is at play.

    6) I’m not going to argue further on Jesus’ historicity. I’m personally going with the consensus view (because I’m not arrogant enough in my ignorance to go against it), and I’ve been led to believe that there’s a more robust debate growing (which may be an echo chamber effect). But I don’t need the argument to defend my position, and will happily discard it.

    7) Again, the natural explanations for a historical Jesus explain his existence just as readily – and with fewer unjustified assumptions (e.g. God brought him into existence for a reason) – as do the theistic ones. It’d be a nice win for the anti-Christian position if someone could prove there *was* no Jesus (an impossible task, and also not a total defeater, I know) – but it isn’t *necessary* because an apocalyptic preacher who got enough followers on board with his story and then got crucified for pissing off the wrong people is perfectly possible. It’s a scenario that’s happened to equal effect more than once – Mohammad being the obvious example. So, again, a historical Jesus provides *zero* support for a divine one, because there’s a perfectly reasonable (and replicable) natural explanation for his story.

    8) “I don’t believe you” simply means that I disagree with your conclusion. It does *not* mean I think you’re wrong, or that I think that some other possibility (“the number is even”) is true. In formal logic, it would be equivalent to saying that your conclusion does not follow from your premises – your conclusion could be right, but you can’t justify it and so you shouldn’t hold to it.

    And I think you’ve rather nailed an important issue here. The atheist position is *all about confidence*. A believer, by definition, has some degree of confidence in their conclusion – otherwise they would reject it. The atheist, on the other hand, has *no confidence in either possibility*, and therefore holds that belief is not appropriate (as always, remember I’m talking about thoughtful people on either side, rather than dogmatic proponents of either position on the existence, or not, of God). And before you turn back to the question of probability, recall that I’ve tried to make it clear that “does God exist?” and “what is the probability that God exists?” are two entirely distinct questions. We can answer “evidence suggests not” *without* making any commitment to “no”.

    This, again, also highlights the burden of proof issue. A theist has a responsibility to justify their claim that God exists, because they should have reasons to support their confidence in that claim. An atheist, on the other hand, has no such responsibility to prove anything about whether God does or does not exist, because their position is that there is no valid reason for confidence either way. Our burden of proof, if you like, is simply to explain why it is reasonable to reject the evidence put forth by theists – and we don’t have to offer any sort of counterevidence in order to do that.

    9) Again, I entirely agree – if I want to convince you that your reasons are flawed, then I need to do more than just tell you I disagree! My point was simply that there is no greater burden of proof on the atheist than to, for example, defend the claim that premise 4 of the teleological argument fails. I don’t have to argue that the conclusion is wrong; I don’t even have to show that the premise is *wrong* (maybe there *is* no physical necessity underlying our universe). All I have to do is show that the premise isn’t justified, which is enough to defeat the argument and therefore to show that the teleological argument, even if it is in fact 100% accurate, is not a justification for belief in God. That is the *greatest* burden of proof that an atheist need adopt, because, going back to my previous point, the atheist position is just that *your confidence in your belief is unjustified, and so you should discard the belief* – so all we need show is why your confidence is misplaced.

    So, I think perhaps points 8 and 9 hit closest to the original topic; but I’ll be interested to hear your responses to whatever catches your attention!

  18. Hello again, no worries about the direction of the discussion. If it’s courteous and constructive, I don’t mind. You make italics by using an i tag as in HTML – the Comment Policy link in the footer tells you how.

    As you said before, this is getting very long, so I am going to confine myself to just a couple of points, and if we can reach some understanding of them, perhaps we can go back to others. OK?

    ” “God did it” isn’t an explanation – it’s much more like a hypothesis masquerading as an explanation. If “God did it” was an explanation, it would actually *tell* us something of value”

    I have seen this comment before, and I really find it strange. The identity of the cause and the process of the cause are both useful pieces of information in any situation. In a criminal investigation, knowing the identity of the criminal is more important than knowing exactly how they did it, though knowing both is better still. Likewise in evolutionary science, knowing that a certain species or feature evolved from another is useful information even if we haven’t yet determined the mechanism, though again in the end we want both.

    So the two are different and both are useful. So how can you say that telling us God created isn’t “useful” information? This is particularly so because that is the topic we are discussing!

    What’s more, naturalism/atheism can’t explain the mechanism by which the universe appeared any more than theism can, and if it ever does theism will have the same explanation.

    So your view, so far at least, cannot explain either the cause or the mechanism, whereas my view can at least explain cause. That must be better.

    ” intelligent beings could exist that bear no resemblance to ourselves”

    I said before, we are not talking about a slightly different universe. We are talking about the vast majority of alternative universes being just composed of Hydrogen atoms, or just Helium atoms, which don’t have the physical and chemical ability to do anything very much. Or a universe that only survives a short time, or which is so spread out that two particles only get together very rarely. You really need to understand how non-life forming most of these alternative universes are.

    “Firstly, you *cannot* claim that there are 10^500 possible universes.”

    I am not claiming that. It is one of the predictions of many cosmologists – if there’s a multiverse, thats’ the sort of number of universes that might form. If you don’t accept that, then you don’t have the one explanation that can be used to avoid design as the cause.

    “That means that any event which occurs more often than, say, 1 in 10^498, is almost certain to occur (if i remember my stats, if you want to catch a 1/x event, you need n = 3x samples to reliably detect it in a trial).”

    Did you know that Roger Penrose has estimated that the odds of getting a low entropy universe like ours needs to be are 1 in 10^10^123? That is way, way, way more than 10^500.

    “I’d be happy to discuss the teleological argument further”

    That’s what I’m doing here. I’m currently reading the latest scientific description of fine-tuning in a book published late in 2016 which I got for Christmas, so I’m fairly up on the science of it all (from a layman’s viewpoint). I think you would benefit from reading up on the science – I suggest you skim through this scientific paper – don’t try to understand all the diagrams and mathematics, but just read the descriptive bits, and you’ll see how strong the evidence for fine-tuning is.

    So I’ll leave it there for now. Over to you.

  19. It seems we’ve entirely strayed from the original topic of discussion, regarding burdens of proof – which is a shame, since it seemed to be going in an interesting direction, and I’d very much like to continue it if you have interest, because I think it’s a core misunderstanding between theists and atheists. Nonetheless, I’m certainly down for a discussion of fine tuning!

    Firstly, though, I think I may have miscommunicated my point about “God did it”. The key there is that it’s not an explanation, it’s a hypothesis. Absolutely, if we could justify the claim that “God did it”, it would certainly be an important thing to know about the origin of the universe, but since there’s insufficient evidence to support the conclusion, it’s nothing more than conjecture and thus *cannot* be considered an explanation.

    Beyond that, though, a fact is not an explanation – it is a description of some feature of reality. If “God did it” is true, that’s then an *important* fact about the nature of reality, but it does not *explain* anything about reality – just as the *fact* that the sky appears blue does not *explain* anything about reality. In order for “God did it” to have any explanatory power at all, you need more facts (such as to how and why he did it as he did) and a theoretical framework to tie them together and provide an explanation of what they mean and allow us to predict about the nature of reality (just like we need facts about scattering of light, wavelength sensitivity of our eyes, mental processing of sensory input, and even how light is generated by the sun in order to explain the sky).

    The non-intervening deistic claim is the worst offender here, because it can offer no information about why god acted as he did, or why he chose this type of universe, or any other useful information. Obviously, different religions offer hypotheses about those issues – but they’re almost entirely post hoc or circular “explanations” of other facts about reality intended to align those facts with the notion of God’s presumed existence (e.g. the mass of an electron is as it is because *that* mass is necessary for our existence, thus God set the mass in order to facilitate our existence – very much a hypothesis in need of testing!). It’s often nothing more than the anthropic principle (we’re here because the constants are as they are, and if they weren’t we wouldn’t be here) in a God-patterned wrapping!

    It is also still subject to objections of special pleading (“God doesn’t need a cause or reason for his actions”) or infinite regress (“what made God?”). Physics doesn’t have this problem, because it doesn’t propose to have an answer – they can say “we don’t know, but we’re doing our best to find out” and leave it at that. The *strongest* claim that physicists can legitimately make is “we believe that the evidence we have points to X” – but *every one of them* knows that finding Y would mean throwing everything out the window (and most of them *actively work* to discover some Y!). “I don’t know” might not be emotionally satisfying (I’d argue against that since, being a scientist [not a physicist!] I generally get quite a bit of pleasure from discovering that there’s something I don’t know and can look into, but I’m probably not a representative sample), but reality doesn’t seem to take our emotional needs into account all that often.

    As to naturalism/atheism can’t explain the mechanism by which the universe appeared any more than theism can and your view … cannot explain either the cause or the mechanism, whereas my view can at least explain cause: the theistic view *proposes* a solution, but in the absence of justification for that conclusion, it is nothing more than that – a proposal. A naturalistic worldview *also* proposes mechanisms, which are *at least* as well supported as the theistic proposals. As far as I can see, the only difference between the theistic and naturalistic models is the confidence of their proponents – “God did it, and he had his reasons for doing so” vs. “well, it looks like there was a big bang, although we don’t know what caused it”.

    I’m going to sort of weave a response from multiple parts of your other highlighted replies, because there’s a lot of overlap in the responses to the objections you’ve raised.

    The first problem I see is the issue of what people claim might be the case vs. what’s possible. It is true that “you *cannot* claim that there are 10^500 possible universes”, regardless of what physics proposes. We have exactly one universe, and cannot currently claim any sort of knowledge about the possibility of anything outside of it. It may be the case that *this* is the only possible universe, and that all the mathematical models which point to a multiverse are entirely wrong. All claims about anything that is not part of our universe – even claims that *it is possible* that there is something outside our universe – are conjecture. Remember, there’s a difference between being able to propose some thing, that thing being possible or impossible, and us *knowing* which is actually the case. There are things which we *know* are impossible (married bachelors); there are things which we *know* are possible (the existence of intelligent beings); and there are many things which are necessarily either possible or impossible, but we don’t know which is actually the case. We also *know* that we can propose all sorts of impossible things, and not *knowing* that a proposal is impossible doesn’t mean it’s possible (or vice versa, for that matter – we can propose that something is impossible, only to later discover that it is, in fact, entirely common!).

    The same applies to Penrose’s calculations. I’ve not got access to the original reference at present (“The Emperor’s New Mind”, if I’m not mistaken?), but I think I’m likely safe in claiming that in order to come up with that number, he has to make a huge number of assumptions about what is possible – and it’s *highly* likely that many of them are not evidentially justified. I’ll gladly retract that claim if proven wrong (and will try to get a hold of the original argument and *own* any error if I can!), but I’ll be amazingly surprised if that happens. At the very least, he needs to assume that other universes are actually possible – which, as I’ve said, is a claim he cannot justify. I’ll also note that Penrose is apparently an atheist himself (and whilst he *does* seem to think the universe has a purpose [a contestable claim in itself], that’s not the same as thinking it was intelligently designed), so if *he’s* not convinced that his arguments support claims for an intelligent designer, why should anyone else be?

    This goes back to my previous post about confidence. The atheistic approach is *all about* lack of confidence in our conclusions about the nature of reality and seeking to develop more reliable ones on the basis of newly discovered facts, whereas the theistic view *starts* from a position of confidence in a particular claim and attempts to explain other facts about reality in light of the conclusion.

    Secondly, the absence of a multiverse does not mean design is the only possible cause. We’ve still got physical necessity and pure random chance, even if there are no other possibilities (e.g. unintelligent design, or “natural selection” of the properties of our universe, as it were – perhaps there was pre-universe “stuff” swirling around, exploding and collapsing, until it happens to hit on a particular type of stable configuration). You mightn’t like those possibilities, but what feels right has little bearing on what is true. I’ve talked a little about physical necessity already, so let’s turn to chance and back to the lottery fallacy. Let’s say we *know* that John Smith won the lottery last week. You would certainly *feel* incredibly lucky if you happened to be John Smith, and the natural question is “well, what are the chances I’d have won?”, but it’s the wrong question, because it’s attempting to define a *known* outcome in terms of the probability of *that* outcome, rather than in terms of the probability of *some* outcome. In Bayesian terms, you’re trying to talk about the magnitude of a post (John *won*) in terms of a prior (what would the chance of John winning have been *before* he actually won?), whilst ignoring a key piece of information (i.e. the thing *happened*!).

    The first thing to recognise here is that we *know* that *something* happened (specifically, that our universe came into existence). It is thus necessarily (tautologically) true that *something had to happen* – specifically, a universe coming into existence. It’s also, I think, fair to assume that the probability of any particular universe is just as likely as the probability of any other universe (and that is precisely what the teleological argument assumes in pointing to the astonishingly small probability of our universe). So, we *know* some universe had to come into existence (because it did), and we both assume that *every* universe is equally unlikely; therefore we would be *equally surprised* by any of them – it’s just that we *happen* to have a universe which allows us to recognise how surprised we should be! It doesn’t matter how small the *prior* probability of our universe is, the *fact* of our universe means that the post probability is 100%, and trying to *explain* it’s existence in terms of prior probability (as opposed to simply *recognising* our good fortune) makes no sense!

    And, going back to previous comments about what we can *propose* about the probability of a universe like ours, versus what is *actually possible* – we can’t even justify the claim that a universe like ours is improbable!

    One final point – “You really need to understand how non-life forming most of these alternative universes are“. No, actually, I don’t. I mean, I *do* understand what is being claimed, but I don’t *need* to in order to reject the argument. My rejection is based on the fact that we *do not know* anything about what types of universes are possible, and we therefore cannot claim that ours is rare or improbable – or common, or even the only type actually possible! All we *do* know is that *our* universe is possible, and since we can’t actually say *anything* about any other *proposed* universe (note, not *possible* universe, since it may be that there *are* no other possible universes), we shouldn’t draw any conclusions about them, or about the likelihood of our own (with or without a creator).


    Anyway, that’s more than enough from me for today. I’ll try to take some time and read through the paper you’ve provided, and see if there’s anything worth digging into with you. I very much look forward to your response to this post, though!

  20. Hi, I want to start by thanking you for the courteous way you are participating in this discussion. I don’t see any reason why people who disagree cannot still behave as friends, but unfortunately many people don’t see it that way. So thanks.

    I don’t think I’ve got a lot more to say about the two arguments about the universe. Yes, “God did it” is a hypothesis, but the point is, it is a hypothesis that seems more probable than the hypothesis that something else caused the universe. So it tells me something about something I want to know.

    I don’t see it as special pleading, it just shows that a God with certain characteristics seems to be a more likely cause than any other cause. Yes, maybe physical necessity or chance is the cause, but at the moment the cosmologists say they’re almost certainly not. Science and common sense work on the idea that we accept the most likely option at the moment, then if something changes, we re-assess. If you want to wait until we know more, then you may be waiting all your life and miss out on the insights that we can get now.

    Obviously Penrose’s calculation isn’t infallible, but it is based on the best theoretical physics we have at the moment. And even if we disregard it, there are scores of other examples of apparent fine-tuning. Again, I can’t see the reason to ignore the best evidence the universe can offer us at the moment.

    I think this statement ”It doesn’t matter how small the *prior* probability of our universe is, the *fact* of our universe means that the post probability is 100%, and trying to *explain* it’s existence in terms of prior probability (as opposed to simply *recognising* our good fortune) makes no sense!” makes no sense. If you were playing poker and your opponent got 5 royal flushes in a row and he tried that explanation on you, I doubt you’d accept it! No, the universe really is amazingly fine-tuned, it couldn’t happen by chance except if there’s a multiverse, and then the multiverse is amazingly fine-tuned and the cosmologists don’t have any “physical necessity” explanation (and even if they did, that explanation would be fine-tuned).

    There really is a question to ponder here.

    ”My rejection is based on the fact that we *do not know* anything about what types of universes are possible, and we therefore cannot claim that ours is rare or improbable”

    This too isn’t the case. Theoretical physics can use well known physical equations to explore all sorts of other universes, and that’s why cosmologists can say that of all the universes that theoretical physics says are possible, only a vanishingly small percentage support life.

    I am currently reading this book which explores all of this (one of the authors wrote the paper I referenced). I recommend you check it out, and check out his blog, to get some more information on the science of fine-tuning.

    I am happy to halt the cosmology discussion here and go back to your interest in the topic of this post.

  21. Hi UnkleE,

    First – thanks for the kind words. I see no value in being antagonistic – nobody learns anything that way! I, too, am enjoying the discussion.

    This is just a short post, and I’ll respond to your previous points tomorrow. But I just wanted to put three questions to you in the meantime:

    1) can you define what you mean by “God”? What are God’s basic properties, etc. if you’ve a page on the topic, feel free to link to it.

    2) how would you distinguish between a universe which is fine tuned for life, and one in which life is fine tuned for the universe it happens to arise in, from *entirely within* the universe (i.e. having absolutely no access to anything which may or may not exist outside the universe)?

    3) I place a tooth under my pillow, and wake to find a coin in it’s place (I’m 5…). Given this, what is the probability that the tooth fairy exists?

    I’ve been reading the article you provided previously, and am finding it interesting (albeit with much skipping of the math – not my forte any more!), although perhaps not for the reasons you might anticipate. I may address it at some stage once I finish with it. In the meantime, however, if you’d like to go back to the original issue of burden of proof, perhaps you’d care to comment on my points 8 and 9 from a few posts back?

  22. Hi again UnkleE,

    At a real computer now, so let me “quickly” respond to your most recent points.

    If you were playing poker and your opponent got 5 royal flushes in a row and he tried that explanation on you, I doubt you’d accept it!
    Firstly, we actually *know* the prior probability of a royal flush (and of five in a row). In contrast, we have *no idea* of the prior probability of a universe like ours (we have hypotheses which *propose* various priors, but since many of those proposals have approximately equal evidential support and offer priors ranging from 1 all the way down to very slightly greater than zero, that doesn’t mean a whole lot). Therefore we can *actually justify* the conclusion that five royal flushes is extraordinary, whilst we cannot do the same for a universe like ours (in fact, on the evidence available, a universe like ours is *the only thing we know is possible* – we *know* that the deck had five royal flushes lined up, and we don’t even know if there are any cards left in the shoe!).

    Secondly, trying to explain an actual event in terms of its prior probability is misguided and nonsensical, because *probability is not an explanation*. The fact that five royal flushes in a row is absurdly unlikely does not in any way explain how or why it happened – fact is, it *did* happen, and the fact that it was unlikely does not preclude it from happening. Likewise, a universe like ours *might* be extraordinarily unlikely, but that tells us *nothing* about why this particular hand was dealt (to stretch the metaphor to breaking point!). Prior probability, at best, tells us that we might want to look for an explanation (because a low prior suggests that an event is surprising and that there *may* be a reason for it), but it doesn’t even mean there *is* an explanation to be had! After all, the prior probability of John Smith winning the lottery is tiny, to the extent that we might consider looking for a reason as to why he won – but we *know* that there is, in fact, no reason other than luck (and the law of large numbers) involved in his winning!

    Again, given that *something* happened (it clearly did), and assuming (reasonably, as discussed in my previous comments) that *every* outcome would be equally surprising (just as a royal flush is *no less likely* than [4H, KH, 2D, AS, 8C] or *any other hand* – it only *seems* that way because we have ascribed importance to the particular combination that we’ve defined as a royal flush; and further, a *series of five royal flushes* is no more likely than *any other pre-specified series of five hands*), then sheer dumb luck *may easily be* the actual (if not emotionally satisfying) explanation for why *this* universe exists, as opposed to any other. Just because we can ask the question, and even propose solutions, doesn’t mean there *is* a solution to be found. Sometimes unlikely things do just happen, and our deciding that there *must* be a reason doesn’t at all imply that there is one.

    Now, we can talk about the *explanatory power* of the varying hypotheses, and the justification for believing that a particular explanation is the correct one. For poker, the obvious hypotheses would be “cheating” versus the null hypothesis, “luck”. Both hypotheses completely explain the event (“the player did something to make it happen” versus “events which are possible can happen”), and given the rarity of the event it’s definitely worth checking for cheating! But in the absence of any evidence of cheating, it doesn’t do any better than the null hypothesis, and should be rejected (because Occam’s razor – cheating requires more assumptions in order to explain the same observations). Ruling out cheating doesn’t prove luck (maybe it was a manufacturing defect in the cards!), but we should always assume the least complicated hypothesis until we have evidence to justify additional assumptions. Same applies for the origin of the universe – a massive degree of luck *does* explain it, and the God assumption simply adds complexity without providing additional explanatory power.

    cosmologists can say that of all the universes that theoretical physics says are possible, only a vanishingly small percentage support life
    Again, this confuses “proposable” with “possible”. Physics is an effort to describe our universe and the facts we observe about it. It is entirely incomplete, with an enormous number of *known unknowns*, let alone the unknown unknowns! And whilst the mathematics describing our universe also make predictions about things outside it, we *know* that our equations are incomplete and thus they cannot justify any sort of epistemological claims about unverifiable things of that type! At best, physicists can say that the best descriptions we have predict certain things – but they cannot claim that those things are even *possible*, unless they can demonstrate that their models are complete and the predictions necessarily true!

    For example, in the article you provided, there’s a lot of talk about how we can change this or that physical constant and end up with a parameter space in which only some minuscule fraction of universes are compatible with life. But that presupposes that the constant in question can actually be varied, or can be varied independently of the other constants, which is *not a justifiable claim*. Take the charge of the electron. Perhaps it *can* vary anywhere between 0.000001 and 1,000,000 (12 orders of magnitude, if I can count). And perhaps the charge of the proton can vary similarly. But even if this variability *is* physically possible (an unjustifiable claim, given current knowledge about reality beyond our universe), what justification is there to assume that the two can vary independently? Perhaps the two are necessarily linked in *every* possible universe, and as such result in a life-compatible combination in all of them. Perhaps that applies to the relationship between *all* constants, such that the number of possible universes *given the 10^12 range available* (which would suggest a space of n(10^12), where n is the number of physical constants in question) is severely constrained by the necessity of the constants being intertwined (maybe to as little as 10^12, if all constants are necessarily linked and a constant can only take a single value, given the value of any other constant in the system). And here’s the critical point – there is no evidence currently or likely available to falsify my argument, or yours. We cannot distinguish between them using the facts available to us, and as such we should not place confidence in either of them. We can find post hoc justifications for either one (the principle of mediocrity might justify the conclusion that constants are entirely variable, whilst the apparent intertwining in our own universe might support that conclusion), but we cannot distinguish between the models they propose unless we look at some other universes (or find a complete description of our own, which necessarily entails one or other conclusion).

    Theoretical physics can use well known physical equations to explore all sorts of other universes
    Again, just to hammer it home – they can *propose* all sorts of universes. That does not entail that those universes are *possible*. I can propose “Samantha the married bachelor” – but that does not mean that she can exist (and, in fact, we *know* that she cannot, because bachelors can be neither married nor female!). Until physics can produce a flawless model of the universe which necessarily entails the flexibility required by the fine tuning argument, or until we can observe all those other life-prohibiting universes with their wacky array of physical constants, (which would actually rather demolish the God hypothesis as a useful explanation for our own universe) then we cannot conclude that they are anything more than married bachelors – that is, things we can propose but which are actually impossible.

    Alright, so I think that covers that. I’ll leave the future direction of the conversation to you – although I would very much like to hear your responses to the questions I posed in my previous post.

  23. Hi, sorry to delay, I’ve had plenty to do and write. Here’s some answers to your questions.

    1. I’m happy to accept the normal definitions of God: supreme being, creator, personal, spirit, etc.

    2. If life somewhat like ours (i.e. natural selection leads to beneficial changes being passed on to more descendants – which doesn’t necessarily have to be so) arises, then it will evolve to suit conditions in that environment. That’s fine-tuning of a sort, I guess, by natural selection.

    But when we talk of fine-tuning in the universe, we are talking about the universe having laws and numbers which allow life to form. Most universes allowed by the laws of physics don’t allow that – either they are too short lived, or they are destructive (mostly black holes), or they are so sparse that atoms would only meet up with other atoms every few million years, or they have such simplified chemistry (e.g. only Hydrogen or only Helium atoms) that nothing we could call life (i.e. reproduction, some ability to do something, etc) could form.

    3. Good question. Let’s assess this by Bayes Theorem, and do it from our perspective and that of the child.

    Let’s guess that the prior probability of the tooth fairy existing would be 1 in a million for us, but 1 in 2 for the child.

    Then we ask these two questions: (i) How likely is the coin if there is a tooth fairy? Both us and the child say 90% (we believe the fairy would be good). (ii) How likely is the coin if there’s no tooth fairy? We say 90% (we believe the parents are good) but the child says 1% ( she doesn’t even imagine the possibility of the parents doing it).

    This is enough information to do the Bayes calculation, and if my maths is right, the answers are:

    We will think the probability of the tooth fairy is 1 in a million (nothing’s changed).
    The child will think the probability is 99% – the data has confirmed her previous uncertain belief.

  24. Hello again, my next comment will be brief. I think all your objections have been considered and here is a very brief summary of the responses to your objections:

    1. Sure we can calculate the probability of a Poker hand more precisely than we can calculate fine-tuning, but the odds against fine tuning are way, way, way larger. It wouldn’t matter how inaccurately we estimate the fine tuning, it is impossibly long odds – much, much more than getting more royal flushes in a row then there are particles in the universe!

    2. Sure “probability is not an explanation”. It is a statistic that requires an explanation.

    3. Yes, “Sometimes unlikely things do just happen”. But when the odds of one aspect of fine tuning are calculated to be 1 in 10^10^123, and almost all the 10^10^123 are the same in that they don’t allow life, it is smart to ask why? Lewis & Barnes (authors of the book I’m reading) say that’s how science is done – investigating unexplained phenomena or statistics.

    4. “this confuses “proposable” with “possible”” – they know of no reason why the other parameter values are not possible. To rule them out, they need a scientific basis, but they currently have found none.

    You really have to recognise, fine-tuning is a real scientific puzzle. The question isn’t whether it’s real, but how we explain it.

    I’m ready to move onto those other matters if you wish.

  25. Hi UnkleE,

    No problems regarding the delay at all. So, to your first post (skipping point 1):

    2) I agree with you that the fine tuning argument is talking about the properties of the universe, but it *is* doing so for the purpose of establishing the possibility of life arising within that universe – so it’s not an unfair question to pose. And I’m afraid I’m not seeing an answer to the question I posed there – how do we distinguish between a universe which is deliberately fine tuned for life, and one which arises entirely by chance and just so happens to permit the arising of life? And, since we’re revisiting it, what *actual facts* about the universe which exists are *not* explained by the “chance” hypothesis, but *are* explained by the “design” hypothesis?

    Note that you *cannot* appeal to prior probability here, because we *know* that P>0 (because our universe exists) and chance can account for *any* P, no matter how small it happens to be – and, in fact, is the most parsimonious explanation of any possible event.

    3) The tooth fairy example highlights a couple of problems. Firstly, if we believe that the presence of a coin tells us *anything* about the tooth fairy, that requires us to presuppose that the tooth fairy exists – which is turning the way we usually investigate things entirely on its head. A naive child, with no prior indoctrination into belief in the tooth fairy, would almost certainly not jump to the idea – they would ask “how did this get here? What happened to my tooth?”. If we give this child reasoning skill beyond his years, we might imagine that the first thing to jump to mind would be “someone’s pranking me”; I suspect magic would be a *long* way down the list of ideas. It’s rather the same with God – even if we assume that apparent fine tuning is what we would expect to see given God did it, there are other ways we can get a universe like ours without having to assume his existence – so he should be a long way down the list of viable explanations.

    I also asked this question to highlight a trap into which we both have fallen in our discourse – and one which is incredibly common elsewhere as well. The probability that the tooth fairy exists *right now* is either 1 or 0 – she either does or does not exist. Speaking about an existential claim in terms of probability makes no sense, because existence is a binary thing – you either exist or you don’t. It might sound like sophistry, but I think it leads to an important confusion in the discussion of issues like this, when we confuse the *fact* of (non-)existence with the *probability that some event would occur given some fact about reality*. Facts are not subject to probabilistic analysis – they *feed into* such analyses.

    When we look at the tooth fairy example, we should *not* ask (as I deliberately did) “what is the probability the tooth fairy exists given this coin?”; rather we should ask “how likely is the coin if we assume that a tooth fairy exists, versus a if we assume it doesn’t?”. This allows us to consider the explanatory power of our hypothesis, which in turn can alter our *confidence* in the truth of our assumption; but it obviously has no bearing on the truth itself. You’ve offering a reasonable analysis of the *confidence* level – the child’s confidence will increase, the parent’s will be unchanged – but you’ve ignored the explanatory power of the hypotheses and the assumptions that they require. The tooth fairy hypothesis requires us to assume that the tooth fairy exists; the “parents did it” hypothesis requires no such assumption, and the *only* way we can think we are justified in believing that the tooth fairy did it is if we lack the imagination to conceive of other, less complex and more evidentially supported possibilities.

    To your second post, now:

    1) You have not – and can not – establish the prior probability of the universe. You *cannot* rule out that our universe is the result of some physical necessity, and thus you cannot rule out that the prior probability of our universe is 1! As the most basic justification, the God hypothesis generally requires that God is a necessary being, in order to avoid the infinite regress problem, so *my* claim to necessity is no less reasonable than the theist’s – and is in fact preferable, since it doesn’t require the assumption of an intelligent being (it simply requires assuming that there is some underlying fact about the universe that we don’t know yet – hardly a stretch!).

    Given that we cannot determine the prior, and that the best ideas we have offer priors ranging from 1 all the way down, we cannot justify the claim that the existence of our universe depends on very long odds. It might be true that God offers a way to get a very high prior – but so do many physical theories which do not require the extraordinary assumption of God. Given our lack of knowledge, we shouldn’t actually assume that *any* of our theories about the existence of the universe are true – but if we’re going to rank them, Occam would have us shuffle any theory which requires God well down the list.

    2) A result does not necessarily require an explanation. It doesn’t even necessarily *have* one. Going back to the lottery – there *is* no useful explanation as to why John Smith won the lottery, and no matter how surprising it might seem, looking for one would be fruitless. Likewise, there may *be* no useful explanation for why our universe happened to be the one to exist – and no matter how hard we look, if our universe *is* just one big fluke, we’re never going to find out why our number happened to come up.

    3) Again, you *cannot* justify the claim that there are 10^10^123 – or even two – possible universes. That is a *hypothesis* which needs evidential support! The five kilogram electron *could* be extant in some other universe – but it could also be a married bachelor, and you have no way to decide (presently, even to investigate) which of those statements is true.

    Certainly, when some observation *seems* surprising – like the existence of *this* universe, when there *seems* to be so many other possibilities – we should most certainly look into it! But we shouldn’t *assume* that there is any actual reason for our surprise, beyond our own ignorance of course (the child is only surprised to find a coin in place of their tooth because they’re ignorant of our tradition of deceiving children in this way), nor should we assume that even genuine surprise is interesting or evidence of any other phenomenon (we should *not* be surprised that someone won the lottery, because someone frequently wins the lottery; and our surprise that John Smith won the lottery is not interesting, because *whoever* happened to win would have been equally surprising from an objective perspective).

    4) “they know of no reason why the other parameter values are not possible. To rule them out, they need a scientific basis, but they currently have found none”. Not *knowing* that something is impossible does not imply that it is possible. In an example such as this, nobody should be saying that the physical constants are or are not variable – logically, they either are or they are not, but *we don’t have evidence to justify either conclusion* and so we shouldn’t accept (believe) either one.

    Like the (a)theist/(a)gnostic divide on belief versus knowledge, this is conflation of fact with belief – those constants *are* or *are not* variable (God *does* or *does not* exist), but the dearth of evidence speaking to either fact means that *we cannot justify a belief either way* and we *shouldn’t pretend to know the answer*.

    The “puzzle” of the fine tuning problem doesn’t strike me as a puzzle so much as an issue of framing. The issue of fine tuning only seems surprising if we ignore the tautological fact that what is possible can happen, no matter how unlikely it might be; and the anthropic principle that if we weren’t here, we wouldn’t be marveling at how equally surprising the high entropy, low gravity, matter-free universe that happened to exist would be. The whole fine tuning argument depends entirely on unjustifiable assumptions in order to explain an event which is *perfectly* explicable by chance; and even if we reject chance (which we cannot), we can improve the prior through assumptions which are incredibly less unreasonable than an intelligent designer. Occam’s razor should dictate that, given far less complex explanations which posses equal explanatory power, we should reject the God hypothesis entirely unless new evidence is forthcoming.

    In fact, the fine tuning argument turns the usual process of knowledge generation (and scientific inquiry) on its head, by *assuming* an explanation (the universe is fine tuned) and asking that we prove the null hypothesis (the universe came about by chance). That’s not how we do things, because the null hypothesis is definitionally unprovable and the simplest of all possible explanations, and not only must the burden of proof lie with the claim that it should be rejected, we should accept the null hypothesis as the best explanation until it fails to account for all of the available evidence (which, in this case, it does not). The fine tuning argument does not possess any greater explanatory power than the null hypothesis – it’s just more emotionally satisfying.

  26. Hi, thanks again.

    ”how do we distinguish between a universe which is deliberately fine tuned for life, and one which arises entirely by chance and just so happens to permit the arising of life?”

    I’m not sure I understand what you are trying to say here. We don’t distinguish between two different universes, we only have one universe. What cosmologists do is say that the probability that this universe arose by chance is exceedingly small, so that few if any think it arose that way except if it is part of a multiverse.

    ”what *actual facts* about the universe which exists are *not* explained by the “chance” hypothesis, but *are* explained by the “design” hypothesis?””

    The incredibly small area of the sample space of possible universes that contains ordered livable universes.

    ”even if we assume that apparent fine tuning is what we would expect to see given God did it, there are other ways we can get a universe like ours without having to assume his existence – so he should be a long way down the list of viable explanations.”

    I have said that there are only 3 basic hypotheses – chance, determined by laws, or designed. Can you please tell me if you know of any other hypotheses that don’t fit into one of these?

    Each hypothesis has its arguments:

    Chance is impossibly small unless we postulate the multiverse. Determined by laws might be right, though (1) the deeper cosmologists go into the laws, the fine-tuning doesn’t go away, and (2) if the laws determine a universe such as ours, we can still legitimately ask why are the laws so fine tuned? Designed explains the cause but not in scientific terms, so scientists prefer to keep looking. But non-believers like Rees and Susskind agree that its a possible explanation.

    ”The probability that the tooth fairy exists *right now* is either 1 or 0 – she either does or does not exist.”

    Look, I am not an expert, but I think you have the wrong idea about probability. Probability is either about our uncertain knowledge of an event, or about an explanation of an event. We can ask what is the probability I’ll pick a card at random and get an Ace, and it is 1/13. If we draw a King, it would be silly to ask what is the probability that this card is a Kin? when we already know. Then the question is What was the probability I’d get this King? and again the answer is 1/13.

    In your tooth fairy example, we don’t know if the tooth fairy exists so we ask the first type of question. And answers like 50/50 or 1 in a million, or zero are all sensible answers.

    ”You have not – and can not – establish the prior probability of the universe. You *cannot* rule out that our universe is the result of some physical necessity, and thus you cannot rule out that the prior probability of our universe is 1!”

    So I don’t think this makes sense. Cosmologists can, with some reasonable evidence, establish the probability of a universe like ours coming about by chance by looking at the sample space of all possible universes allowed by the laws of physics, and seeing what percentage of that sample space is “like our universe”, e.g. in the sense of allowing intelligent life. They have done that, and the odds are vanishingly small unless there is a multiverse. You are right that we cannot rule out physical necessity, but that simply means that it couldn’t have been different because of certain reasons, and then we can ask how likely is it that it had to be that way. Much harder I imagine to calculate, but still sensible.

    ”A result does not necessarily require an explanation. It doesn’t even necessarily *have* one.”

    No of course not. But science is all about finding answers and explanations. And so is a quest to know if God exists. All this fine tuning information is helpful to both searches.

    ”the fine tuning argument turns the usual process of knowledge generation (and scientific inquiry) on its head, by *assuming* an explanation (the universe is fine tuned) and asking that we prove the null hypothesis (the universe came about by chance).”

    This isn’t true either. Scientific fine tuning is simply the true observation that, of all the universes allowed by theoretical physics, a vanishingly small number would allow life. That is a fact, no doubt about it. Then we look for explanations for that fact. The scientists have looked long and hard and the only scientific explanation they can find is the multiverse. The fine-tuning argument is a philosophical argument based on the fact of fine-tuning that argues that the universe was designed, but arguing that neither chance nor necessity are at all likely. It’s nothing to do with the argument being emotionally satisfying, it is to do with it being correct, and if anyone thinks its not correct, it is best for them to demonstrate which premise is wrong.

    I really think we should close up this discussion of fine tuning here. I think you need to read more about it. I strongly recommend Lewis and Barnes A Fortunate Universe, which I have just finished reading. It is deep at times, but well worth it.

    Do you want to go back to the other matters you were interested in?

  27. Hi UnkleE,

    The problem with going back to a discussion about the burden of proof is that we’re sort of already having it. It’s complicated by all the details around fine tuning (but it’s also hard to leave those misunderstandings just lying there, too). However, let me pick a couple of key points out of your last post and try to turn them back to that discussion.

    “how do we distinguish between a universe which is deliberately fine tuned for life, and one which arises entirely by chance and just so happens to permit the arising of life?”

    Your confusion here goes right to the heart of the principles underlying hypothesis testing. Effectively, when we conceive a new hypothesis to explain some observation, we imagine two versions of our world – one in which the hypothesis is true, and one in which it is false. We then try to imagine what would be *different* in those two worlds, so that we can try to figure out which one is ours.

    Consider an example. I have a hypothesis, “cheomitine is an effective treatment for pain”, and I can imagine a world in which this hypothesis is true, and one in which it isn’t. In the former, I can reasonably assume that a clinical trial would show a significant difference in pain between people treated with cheomitine versus those given a placebo; in the latter, I would expect no difference between those groups. I can then construct an experiment to figure out which of those worlds happens to be the one we live in. But prior to performing the experiment, I *cannot* say that cheomitine does or does not work – all I can say is that the best evidence we have to date does not show that cheomitine is an effective painkiller.

    Then, when we’re planning the trial, we need to figure out how to design it. For various reasons which make “proof” in most science difficult or impossible, we don’t try to *prove* that cheomitine works; instead, we set out to prove that the most reasonable pre-trial assumption – cheomitine does nothing, which is reasonable to assume because *most* things do nothing *and* we have no evidence that cheomitine does anything – doesn’t adequately describe the real world. In other words, the null hypothesis (the most parsimonious explanation for all the facts we have available to us) is that cheomitine is biologically inert, and for practical (as opposed to epistemological) purposes, we should assume that that’s true until we have evidence to the contrary (after all, we shouldn’t be wasting patients’ time and money on something that is *probably* not going to help them).

    And finally, when we look at the results, we can draw some tentative conclusions about the world in which we live. If the results show a positive difference between treatment and placebo, it’s reasonable to conclude that the null hypothesis *does not* adequately describe the world in which we live – because it would not predict such a difference. Of course, we’d want to replicate our findings, check out different patient populations, and so on, in order to increase our level of confidence in the results – but we at least have evidence to call the null hypothesis into question. But, if the results demonstrate no difference, we’ve found evidence which both rejects the effectiveness hypothesis *and* supports the null hypothesis. Again, we might want to replicate the findings to be more confident, but we should most definitely not reject the null hypothesis.

    So, if we now turn this process onto the teleological question, how do we figure out which hypothetical world – one that is fine tuned for life, versus one which *permits* life by chance – we live in? What facts about reality are incompatible with the “chance” hypothesis but accounted for by “God did it”? What does the theistic model predict that the “chance” hypothesis does not? My opinion, I hope it is obvious, is that we *cannot* determine which of them we live in on the basis of any available evidence (and *speculation* about the prior probability does not constitute evidence about what the prior actually is; and even then, a low prior does not *at all* exclude chance as an explanation), and we should therefore not express any confidence or belief in either conclusion. For practical purposes (if there are any) we should default to the null hypothesis (compatible with, not tuned for, life). We should simply say “I don’t know”, and whilst it’s absolutely worthwhile to continue our investigations into the origins of our universe, and to discover if there *is* anything more to the compatibility of our universe with life like ours than simple chance, we shouldn’t believe that there is anything more than chance at play until we have evidence which demonstrates that chance *cannot* be the reason.

    I also must continue to dispute the very foundation of the fine tuning argument, in that I do not accept the claim that the universe is fine tuned for life (again, I don’t *deny* it, but I don’t *accept* it, either – consistent with the view that we should never adopt *any* belief unless doing so is justified by an appropriate amount of evidence). Firstly, the notion that life is fine tuned to the universe (which can occur through natural selection) is an entirely parsimonious explanation for the existence of life in any universe which happens to permit it. As to why the universe happens to permit it – we don’t know, and we have no epistemologically valid grounds for drawing any sort of conclusion, so we shouldn’t. We have the *propositions* put forth by theoretical physics (and religions) but, despite this being a rather trite point, there’s a *reason* it’s called “theoretical”. Theoretical physics proposes all manner of strange and, to the average person, unimaginable things, from a purely mathematical universe, a holographic universe, a near-infinite multiverse, a physically necessary universe, a universe which divides every time a “choice” is made, and so on. There is *no* basis for selecting between any of the available options with any degree of confidence, and thus we shouldn’t subscribe to *any* of them – nor should we subscribe to “God did it”, for the same reason.

    One other point, about the tooth fairy. Probability describes the likelyhood of a possible event coming to pass. If an event occurs, though, the probability that it happened is 1 – it’s a tautology. When applied to an existential claim, though, speaking of probability makes no sense – the probability of something coming *into* existence might be 0.5, but the probability *that* something exists (i.e. came into existence previously) is 1 or 0. Right now, as I write this, the tooth fairy *does* or *does not* exist – she doesn’t exist to an extent which depends upon how likely we think it is that she exists (otherwise we need to commit to multiple or entirely subjective realities, in which the tooth fairy exists for the child, but does not for the parent! Or worse, we have to break the fundamental logical principle of excluded middle (“X” or “not X”) and posit that she *partly* exists!).

    When we talk about hypotheses, though, we talk about *explanatory power*, not the probability that they’re true (actually, lots of people *do* talk about that, but they’re fundamentally wrong). What we actually (should) talk about in that context are things like “what is the probability that I would get this result if the null hypothesis is true?” (which is very different to “what is the probability that my hypothesis is true/false?”). Or, even more commonly, we talk about the *confidence* we have in an explanation given the evidence available – I’m 99% confident that the theory of evolution provides a good description of the basic processes underlying evolution and speciation, given all the available evidence to support it; I’m 0% certain that a god does or does not exist; and I’m say 75% confident that Bigfoot does *not* exist, because of the lack of evidence that has been found despite extensive searching. In simple terms, I’d be *extraordinarily* surprised if the theory of evolution were overturned; I’d be not at all surprised if a (deistic) god exists, and I’d be fairly surprised if someone found a Bigfoot tomorrow (not a previously unidentified species of mammal, mind, but a *Bigfoot*, at least close to what is described in popular culture). We could extend that to a specific god – let’s say I’m 25% confident (a number I won’t try to defend, since it’s somewhat random) that the Christian god *doesn’t* exist (because problem of evil, divine hiddenness, and so on), so I’d be reasonably surprised if sword-mouth Jesus came down from the sky tomorrow to kick off the prophecies described in Revelation.

    As soon as we start confusing confidence in our explanations with “the chance that X exists”, though, we start using very muddy language which confuses the validity of the hypothesis we’re proposing or our confidence in our beliefs with the underlying facts about reality. I think we’ve both been guilty of that type of confusion in this very discussion, and I think it’s worth pointing it out and focusing on, because that confusion between what is true and what we believe seems to keep coming up as a sticking point in a/theist discussions.

    So, all this hopefully ties back to the burden of proof issue that I initially raised. The basic epistemological position on *any* issue is that we should never believe *any* claim is true unless it is justified by sufficient evidence (“sufficient” depending on the claim in question). That also means we shouldn’t believe that it’s false, either – we should remain both agnostic (not claiming knowledge) and uncommitted (not holding a belief). When looking specifically at questions of gods, we move to the atheist position, which is merely a specific application of that epistemological principle – there is insufficient evidence to justify a belief in the existence or non-existence of god, therefore we should not adopt a belief regarding the question until better evidence is forthcoming. There is *no* burden of proof attached to the atheistic position on the existence of a god, because it is not *actually* a claim – it is simply the statement that we need to wait *before making a claim* about existence!

    When dealing with a theistic position, the burden of proof necessarily lies with the theist, because the theist must argue that the null hypothesis should be rejected, and needs to provide evidence to justify their claim. In arguing *against* that position, the atheist need at most demonstrate why the evidence presented is insufficient to justify rejecting the null hypothesis. For example, the theist must demonstrate that our universe *is* in fact fine tuned for life (which means finding some fact about our universe which chance alone *cannot* explain); the atheist, on the other hand, need simply point out that there is no way for us to actually determine whether our universe is tuned for life, or whether life arose in a universe which happened to permit it. The theist must demonstrate that 10^10^123 (or even two) different universes are *possible* (not just *imaginable* or even calculable under some assumptions about “extra-universal” laws of physics to which we currently have no access) in order to argue that there even *is* fine tuning to be explained; the atheist simply needs to point out that we cannot empirically or rationally demonstrate that anything outside our universe can exist, so we shouldn’t believe it to be the case and we shouldn’t base a claim of fine tuning on that assumption. The atheist *certainly* doesn’t need to commit to any alternative explanation – they simply need to argue that the theist’s conclusion is not justified by the facts that they have available, and so should be ignored until such time as sufficient supporting evidence is forthcoming.

    Now, to be fair, some atheists *do* take it further for specific gods – “the problem of evil is evidence against the existence of a benevolent god, therefore we are more justified in believing that such a being does *not* exist than believing that one does” – and in that case, the atheist adopts a burden of proof, in that they need to justify their claims and demonstrate that, say, evil is logically incompatible with a benevolent god. But no thoughtful atheist I’ve ever come across would argue that all gods are *impossible* (just like you can’t rule out solipsism, you can’t rule out deism – an otherwise non-intervening creator god).

    For practical purposes, of course, we often need to act in the absence of epistemologically valid justification – but in those cases, we *always* default to the null hypothesis, because it requires the least number of unjustified assumptions. That is not to say we should *believe* that the null hypothesis is correct – we shouldn’t – but when necessary we should *act as if it is* until we have sufficient justification for doing otherwise. For cheomitine, we assume that it doesn’t work – because *most* things don’t work – and we don’t use it to treat patients (no matter how desperate their need) until we have evidence that it does. For god, we assume he doesn’t exist – because we have no positive evidence that he does, and positing existence of a supreme being is a much larger assumption than positing any non-intelligent process which might lead to a universe like ours, no matter *how* remote such a possibility might seem from our obscenely limited perspective – and don’t waste our time in worship. Or for the issue of fine tuning, we assume that the universe is *not* fine tuned for life because, again, there is no positive evidence to dispute that (and the untested – perhaps untestable – fairy tales dreamed up by theoretical physicists and religious prophets do not in any way constitute evidence), and so we don’t base conclusions on the assumption that it *is* fine tuned. Again, though, no burden of proof is required to justify such a basic principle of epistemology as withholding judgement philosophically, whilst acting in accordance with the most parsimonious explanation if required to do so.

    So, I’d most definitely like to hear your response to the question about distinguishing between a fine tuned and a life-permitting universe, if you have any thoughts on the matter; otherwise, I as always look forward to your reply!

  28. Hi Cheomit, and welcome again.

    ”Your confusion here goes right to the heart of the principles underlying hypothesis testing. …. What facts about reality are incompatible with the “chance” hypothesis but accounted for by “God did it”? What does the theistic model predict that the “chance” hypothesis does not?”

    I don’t think there is any confusion about this. The chance model is impossibly improbable. The estimate for the probability of the universe starting in such a low entropy state by chance is 1 in 10^10^123. I don’t think anyone could be that certain that God doesn’t exist, so the probability of God MUST be greater than those odds.

    So I don’t think we have to say “I don’t know”, we can say God is way more probable than chance – except if there’s a multiverse.

    ”There is *no* basis for selecting between any of the available options with any degree of confidence”

    But there is a strong basis – the very long odds of it happening by chance, unless there is a multiverse.

    ”What we actually (should) talk about in that context are things like “what is the probability that I would get this result if the null hypothesis is true?” (which is very different to “what is the probability that my hypothesis is true/false?”).”

    But that is exactly what the fine-tuning argument does. It says, the hypothesis that the universe arose by chance and there is no multiverse, has almost zero probability..

    ”there is insufficient evidence to justify a belief in the existence or non-existence of god, therefore we should not adopt a belief regarding the question until better evidence is forthcoming”

    I have no problem with this (in theory, at any rate). But I wonder, do you visit theistic and atheistic websites in equal numbers and argue equally strongly against both viewpoints?

    That isn’t a rhetorical question, I’d really like to know.

    ”For practical purposes, of course, we often need to act in the absence of epistemologically valid justification – but in those cases, we *always* default to the null hypothesis, because it requires the least number of unjustified assumptions.”

    Always? I can think of cases where that isn’t at all the obvious way to go.

    What if the risk of one course is much larger than the risk of the alternative? It would then make sense to reduce risk. Say you are in the jungle and you think you hear the sound of a lion. But you are not sure. So do you do nothing, or do you reduce your risk and take the appropriate action? The answer is obvious.

    This is in fact arguably the case with God. If there’s no God, then your life and mine are both ultimately meaningless and we both have as much chance of being happy as the other. (Actually, studies show the religious people have a slightly greater chance of being happy, but let’s ignore that.) But if there’s a God, then it is just possible that believing in him and doing what you think he wants may be a safer option than ignoring him. So the risk factor makes it smarter to lean on the side of believing or obeying. (I know this simplifies the issue, but the principle is clear.)

    ”For god, we assume he doesn’t exist – because we have no positive evidence that he does, and positing existence of a supreme being is a much larger assumption than positing any non-intelligent process which might lead to a universe like ours, no matter *how* remote such a possibility might seem from our obscenely limited perspective”

    But this is the point! We do have evidence, and fine tuning is part of it. The possibility that it happened by chance is pretty low unless there’s a multiverse, so there is evidence, even if not totally conclusive evidence, that God exists. So I suggest that thinking it happened by chance is a larger and stranger assumption than thinking it was designed.

    ”the question about distinguishing between a fine tuned and a life-permitting universe”

    These are really the same thing. The phrase “fine-tuned” is just a short-hand for saying a universe like ours is very improbable by chance. We can express that using various facts – the probability that the universe would last more than a particular time, the probability that chemistry more complex than a single element would appear, the probability that stars could form, and the probability of intelligent life, or just life.

  29. Hi UnkleE,

    Thanks as always for the interesting reply. I’m really going to jump around here, because there’s a lot of intertwined issues to deal with, so I’m sorry about that.

    do you visit theistic and atheistic websites in equal numbers and argue equally strongly against both viewpoints?

    Realistically, I rarely engage in theistic discussions. I can’t even recall, now, why your post happened to catch my attention! However, I’m quite confident in saying that I tend to focus on arguing against poor epistemology *whenever* I engage in an argument, on or offline – partly because it’s something I care and am informed about and that I think underlies many problems with people’s understanding of all sorts of issues, but also because I’m reasonably well aware of my own limitations and generally try to avoid taking a firm stance (or to defend my stances) on issues about which I might not be reasonably informed. I’d much rather probe the *reasons* a person has for their beliefs, than to defend my own or defeat theirs – if nothing else, I find I learn more by focusing on why, rather than what, a person believes.

    So, whilst I don’t routinely engage in arguments with people arguing *against* the existence of god (as opposed to arguing against *belief* in god – which I don’t do at all often either), I have done and would absolutely do so again, at least as vigorously as I have here. I tend to argue about the epistemological underpinnings of arguments, on all sorts of topics (politics, gun control, free speech, religion and whatever else grabs my attention at the time) rather than their conclusions, and atheists can be just as epistemologically challenged as theists!

    we *always* default to the null hypothesis leading to you are in the jungle and you think you hear the sound of a lion

    You’re making a false analogy here, because you’ve failed to identify the null hypothesis. The null hypothesis, upon entering the jungle, is “there are no predators in the immediate vicinity”. But, we know that predators of various types live in wild environments; a sound that could be a predator is evidence that a predator might be immediately nearby; therefore, it is both practical *and* reasonable to assume that there’s an immediate threat and to act accordingly (which may mean running screaming; but it would more likely mean raising your guard and seeking *some* protection whilst you investigate further, rather than abandoning whatever purpose you had being there despite the known risk of predators in the first place).

    Note, however, that this is not the same as assuming there’s a predator nearby just because you’re in the jungle, without the additional evidence of the relevant noise – the *noise* is the evidence which leads to your rejection of the null hypothesis (“no predators are nearby”), and even if it’s poor evidence or turns out to be the wind, you’ve at least got *some* basis for rejecting it (because the noise seems inconsistent with the null hypothesis – even if it turns out not to be and you’re forced to readopt it). What such evidence is there to support rejecting the “no god exists” null hypothesis, let alone the “the Christian god does not exist” null hypothesis? Or, in the fine tuning argument, what evidence is there to justify the belief that because you exist, there must have been an intervention in the creation of the universe to ensure that you would exist, over the null hypothesis that you exist by the benefit of nothing more than good luck?

    the question about distinguishing between a fine tuned and a life-permitting universe

    First off, these are in fact very much *not* the same thing. A [universe which is specifically designed for life] and a [universe which is simply compatible with life] are different in their *cause*, which is the whole issue at question – it’s the *cause* that we’re trying to establish here! But if you cannot distinguish between them based on any information available to you from within the universe (the situation we find ourselves in right now), then what basis do you have for the claim that it was tuned, as opposed to simply accepting the *fact* that it is compatible (without claiming to know *why* – or even to know if “why?” is a sensible question!)? That’s the purpose of the thought experiment – to figure out what would differ between two worlds which came about in different ways (one by design, on by chance) and to figure out what, if anything, would differ between them. If you can’t conceive of any sort of difference, let alone find evidence of one, why would you assume the more complex hypothesis to be true?

    Let’s go back and break down the construction of the fine tuning hypothesis. Here are the relevant, *indisputable* facts that we have about the universe:

    1) The universe exists;
    2) The universe is compatible with the existence of life.

    Also, just to be clear – I’m going to talk about things like “before the universe existed”, which assumes time has some meaning outside our own universe. I’d like to be clear that I’m doing so for convenience, and am not making any sort of metaphysical claim!

    Now, it may or may not be true that there is an explanation for why the universe we inhabit permits the existence of life. The null hypothesis would simply be “chance”, because:

    1) The universe exists;
    2) Before the universe came into existence, it was therefore possible that the universe could exist;
    3) Any event that *can* happen can happen by chance;
    4) Therefore, the universe could have come into existence by chance.

    Now, the null hypothesis is, as always, boring and uninteresting, because it doesn’t really tell us anything new or particularly useful – if the universe came about by chance, then that’s the end of the story. And the null hypothesis is not always right, either! So, we should absolutely look into it.

    That’s where “fine tuning” comes in – as a proposed explanation for the *fact* of compatibility and the *assumption* of surprisingness:

    1) The universe which exists is compatible with the existence of life;
    2) The prior probability of a universe which permits the existence of life is infinitesimal;
    3) It is possible that some process which results in fine tuning of the universe might be able to account for the low prior probability;
    4) Therefore, the universe may be fine tuned to allow life to exist.

    In order to justify (4) (which is the *starting point* for the teleological argument), you need to prove (2), and you simply cannot do that – we have no access to anything outside our universe and no justification for claims about what other universes are or are not possible. As I said previously, theoretical physics is a bunch of math-filled fairy tales when it comes to any sort of guesses about anything that may or may not exist outside our universe – *many* of which could slot with equal validity into (3) (and thus the conclusion). The fact that the math does a good (but hugely imperfect) job of predicting *our* universe doesn’t mean that the many assumptions physicists might make about anything else outside it are valid. Physics proposes *all sorts* of universes and realities – far more than just the multiverse that you appear to be most focused on – which could *all* account for the existence of our universe, but just like “God did it”, they’re all guesses and we shouldn’t believe *any* of them are actually true until we have evidence to support them. Since we currently have *no* access to anything outside our universe, and thus have *no* way of testing these myriad hypotheses, we should simply accept that we don’t – and can’t – know, and should *keep looking* for the real answer, even if we can never actually get to it, instead of pretending that we have one, or that we have reason to think that we do.

    In short, though, We do have evidence, and fine tuning is part of it is a false claim, because “fine tuning” is an *unproven hypothesis* about the reason why the universe we inhabit is compatible with life – it is not *evidence* of anything whatsoever. We cannot assume that the compatibility of the universe with life is evidence of fine tuning in order to prove that the universe is fine tuned! There may be no reason (chance or physical necessity); there may be one of very many reasons (physical limits on the variability of the constants which limits the types of universes possible, a multiverse, “natural selection” of universes, God, etc.) – but we cannot distinguish between them with the evidence we have available, and so we should talk about *the compatibility of the universe with life*, not the “fine tuning” of the universe – because, again, fine tuning is a hypothesis *attempting to explain that compatibility*.

    thinking it happened by chance is a larger and stranger assumption than thinking it was designed

    This is not true, either. In order to accept that it happened by chance, we only need facts – something that did happen could have happened, something that can happen can happen by chance, and the universe *could* have happened because it did; therefore, the universe could have happened by chance. In contrast, for “God did it” you need to posit the existence of an intelligent being, entirely unlike anything that we have any sort of example of outside fiction (and not even then, if we’re to believe many theists), who not only created the specific universe we inhabit over all the other similar universes he could have chosen, but is necessarily greater and more complex than it is. The God hypothesis relies on a *lot* of assumptions.

    Back, then, to hypothesis testing. We have a null hypothesis – the universe came into existence by chance. We don’t *believe* that this is true (because the null hypothesis is a theory of exclusion and can only be proven true by excluding *all* other known and unknown possibilities), but for practical purposes it is the most reasonable assumption because it doesn’t *require* any assumptions itself. At the same time, we recognise that the universe *appears* to be surprising – theoretical physics *does* propose all sorts of models which would render our universe horrifically unlikely. So, we go looking for a reason as to why *our* universe happens to exist. You propose “God fine tuned the universe to our needs”; physicists propose multiverses and holograms and all manner of other ideas. But all of these proposals are hypotheses in need of testing – they are *not* explanations.

    When we examine these myriad hypotheses, then, what we need to do is look at the universe and find some facts which are incompatible with the null hypothesis, but are predicted by one of the alternatives. Firstly, there is currently *no* fact about the universe that is incompatible with the existence of the universe by chance – the fact that it is compatible with life is not incompatible with the possibility that it came about by chance, and it’s not *surprising* because if chance is at play, then *every* universe would have been equally surprising (and therefore *none* are truly surprising), assuming there were any intelligent beings around to recognise the uniqueness of the universe in question.

    But we can at least make some predictions. Under a fine tuning model, we’d expect some evidence that the constants are fine tuned rather than luck – perhaps a model of physics, which completely describes *everything* in the universe but which depends upon a vastly different mass of the electron, suggesting that the mass of the electron was tampered with. We’d expect there to be evidence that other universes are possible (which requires more than theoretical physicists theorising about what numbers on their page are in reality variable!) – perhaps through direct access to one, or through another *complete* physical theory that depends on that assumption. Without evidence of that sort, we can’t exclude the null hypothesis, and we shouldn’t adopt a different hypothesis (because even if the null hypothesis is wrong, *we’re very probably going to be just as wrong* if we choose an alternative essentially at random from the many available to us – let alone the ones nobody has even thought of yet!).

    we can say God is way more probable than chance – except if there’s a multiverse

    No, we can’t. We might say god is a mechanism which might overcome the low prior probability of our universe, meaning that our universe is less surprising if god exists. We might say a multiverse is a mechanism which might overcome the low prior probability of our universe, meaning that our universe is less surprising if a multiverse exists. We might say that some physical constraints on what universes can exist is a mechanism which might overcome the low prior probability of our universe, meaning that our universe is less surprising if such constraints exist. We *can’t* say that *any* of those mechanisms are actually real, let alone which one! By your own repeated admission (“almost zero” chance), it remains possible that the universe truly is entirely and unreasonably “surprising” (viewed through the lens of the lottery fallacy), our discomfort with that possibility having no bearing on that. In the absence of any evidence to demonstrate that chance is *impossible* (or, more specifically, that the null hypothesis fails to account for), we shouldn’t *practically* – let alone epistemologically – reject it in favour of our own preferred religious or theoretical physics-based fairy tale.

    We can say “God/multiverse/whatever are explanations which would *adjust* the extremely remote prior probability of our universe in our favour” (“remote prior probability”, again, still being an unjustified assumption in itself), but we *cannot* say that one is more probable than the others. What we *want* is a hypothesis with better *explanatory power* – which means, a hypothesis which explains the facts we have about reality better than the null hypothesis. But given that the null hypothesis does not fail to account for any *known* (not *proposed*) facts about the universe, we cannot justify rejecting it or investing our belief some other explanation.

    Now, to a few small points:

    The estimate for the probability of the universe starting in such a low entropy state by chance is 1 in 10^10^123. I don’t think anyone could be that certain that God doesn’t exist, so the probability of God MUST be greater than those odds.

    That estimate is entirely theoretical, and so cannot be the basis for any claims about what *is* true – only what *might* be. Furthermore, the probability of God is 1 or 0 – there is no other option. The *confidence* we could have in our universe coming to exist (prior to it doing so) would be higher if God exists, certainly; but that has no bearing on *whether* he exists, nor does the *fact* that our universe exists give us *any* information about the prior probability of it coming to be (and it is *still* entirely consistent with chance or any number of other mechanisms).

    The *only* way for “God” to be more probable than *any other explanation* is if God exists (i.e. p = 1). By definition, then, all other explanations would have p = 0 (because “God did it” is what *happened*, therefore none of the other things did!). The God *hypothesis*, though, certainly has greater explanatory power than the null hypothesis *if and only if* the universe is fine tuned (which we cannot establish as being true – fine tuning, again, is a *hypothesis* to explain the apparent surprisingness of a universe which is compatible with life, it is not evidence!). Of course, even *if* the universe is fine tuned, that doesn’t at all prove God – a multiverse or a “natural selection” of universes, both arising through purely physical processes (even if they have to be accepted as necessary facts – because God is typically put forth as a necessary being) would remain far more reasonable hypotheses than a god.

    [The fine tuning argument says] the hypothesis that the universe arose by chance and there is no multiverse, has almost zero probability

    No, the fine tuning argument *relies on the assumption* that the prior probability of the universe coming to exist is almost (but not actually) zero. That low prior is an *assumption*, it is not a known fact – it’s barely a justifiable claim, given that the *only* thing supporting it is theoretical physics (which also supports the disconfirming multiverse, as well as myriad other hypotheses which would negate any surprise at our universe existing by chance without appeal to God).

    So the risk factor makes it smarter to lean on the side of believing or obeying.

    I have to say, I’m a little saddened to see you bring Pascal’s (bankrupt) wager into the conversation. Pascal’s wager is yet another argument which puts the conclusion ahead of the evidence and assumes that the god that *you* happen to believe in is the *only* god on the table. The obvious first place to point is at what I suspect is a fact, that neither you nor I have any particular concerns about the hell described in the Quran – but, by the logic of the wager, shouldn’t we both be worried about the seemingly far more vengeful Allah? Even *within* Christianity there are so many hypotheses regarding the consequences of non-belief – for example, some Christians are convinced I (and perhaps you) will burn eternally; some think it’ll be just for a while; some think there *is* no hell; and some think that “hell” is just “distance from God” – which, for an atheist, isn’t a problem at all, because the *moment* I possess sufficient evidence for God’s existence (and I think being in his presence might get that done), I’ll be convinced and will absolutely believe in him.

    When you get right down to it and dispense with the assumption that the theist using the argument at least has the *most likely* god, Pascal’s wager is more an argument *for* atheism than against it. After all, if Allah is the one true god, who’s he going to be more pissed at and vengeful toward – me, who simply doesn’t believe in any god, or you who worships a false idol?

    OK, so that was one horrifyingly long post… Hopefully some weekend reading for you to consider! All the best, UnkleE – I look forward to hearing your responses.

  30. Hi UnkleE,

    Sorry to double post again, but one more important question occurred to me.

    How do you reach the conclusion that a universe like ours is horrifyingly unlikely by chance? So far as I can tell, it must be physics and/or intuition; please let me know if I’ve overlooked something, and what the option (or mix) for you personally is.

  31. Hi Cheomit, that is indeed a very long comment! (More than 3,000 words, and thus longer than the original post!) I will try to be brief by limiting the matters I address.

    Re my jungle analogy. You have addressed all sorts of matters here, except the point I made – that sometimes risk changes how we assess matters.

    ”That’s where “fine tuning” comes in – as a proposed explanation for the *fact* of compatibility and the *assumption* of surprisingness:“

    I think this is a misunderstanding. When I talk about “scientific fine tuning”, I am talking about scientific facts, not an explanation of those facts. The two things are very clear, and I have been consistent (I think) in my terminology and definitions. Please read this clearly.

    1. Scientific fine-tuning is the scientific fact that, of all the possible universes allowed by theoretical physics, only a vanishingly small number allow life (or even structure and chemical matter) to form. It says nothing about design or any other explanation.

    2. The theistic teleological argument starts with this fact and argues that there are logically only three explanations for these facts – random or non random, which can be divided into designed and necessary by logic or laws.

    I think a lot of what you are saying is assuming the scientists and I are talking about 2 when we are talking about 1. Only when you can understand 1 can we then discuss 2.

    Re Null hypothesis. The null hypothesis is that this universe came into existence by chance. The alternate hypothesis is that it didn’t. We test hypotheses using statistics. As Wikipedia says: ”If the data-set of a randomly selected representative sample is very unlikely relative to the null hypothesis (defined as being part of a class of sets of data that only rarely will be observed), the experimenter rejects the null hypothesis concluding it (probably) is false.”

    In the case of fine tuning, calculations have the shown following odds if the universe was generated randomly:

    Penrose: the probability of starting with such a low entropy universe = 1 in 10^10^123. Smolin: the odds of stars forming = 1 in 10^229. Leonard Susskind: “To make the first 119 decimal places of the vacuum energy zero is most certainly no accident.” There are many other facts with similar long odds.

    These odds are way way beyond what is necessary to reject the null hypothesis. The only way, the cosmologists say, we can accept chance is if we postulate a multiverse generating so many universes that it becomes less improbable that our “fine-tuned” universe would arise. I really don’t think we can go further in this discussion until you understand this and we can agree on it.

    ”That estimate is entirely theoretical, and so cannot be the basis for any claims about what *is* true – only what *might* be.”

    So are you saying that most of modern Physics is wrong? We have put people on the moon, sent space probes to other planets, developed scientific theories and laws about particle physics, cosmology, etc, that have been verified by experiments, but you are saying it’s all wrong?

    My response to your second post: it is a finding of modern cosmology. Read the Barnes paper I referenced before. Read my summary here (to which I now have to add some some extra information since reading Lewis & Barnes). There are very few cosmologists who haven’t concluded that way.

    Finally, I didn’t mention Pascal’s wager, I talked about risk = consequence x likelihood. Do you think it is wrong to think of risk?

    I think that is enough for now. Thanks.

  32. Hi UnkleE,

    We’ve covered some substantial ground beyond the original post, so I suppose it’s not unreasonable that things might expand!

    I didn’t mention Pascal’s wager, I talked about risk = consequence x likelihood.

    You might not have mentioned Pascal’s wager by name, but that’s the argument you’re invoking when you comment on the risk of not believing. Regardless, any argument from “consider the risks” merits the same response – why should an atheist be concerned about *your* theology’s consequences when *you* are not concerned about, say, Islamic hell? Why does the risk vs. consequence analysis *only* take your god, and not all gods, into account? And why would a god, no matter how vengeful, be *worse* to atheists attempting to be rational and to apportion their beliefs to the best available evidence than to those who worship false idols in ways which scriptures tell us would anger the various gods? Not saying Allah *would* treat me better than you – but odds are he’d not treat me *worse*, and it’s *possible* he’d treat me a little better!

    sometimes risk changes how we assess matters

    Indeed it does, but it doesn’t change the validity of existential claims – particularly when one of those claims is about the existence of the risk itself, especially when responding to it requires dismissing equal (or even greater) but competing risks! And indeed, certain risks might change our *practical* response to a claim (e.g. unconfirmed local reports of a rabid wolf in a region known for a large wolf population), but it should *never* affect our epistemological stance (the bear exists or it does not – and whilst we might carry our gun a little more readily out in the woods, we shouldn’t pretend to *know* that the wolf is out there; or, if we shoot a rabid wolf, we shouldn’t pretend to *know* that it was “the one”). But when we have multiple comparable, competing and indistinguishable risks (Christian vs. Islamic hells), we *definitely* cannot justify even a practical-level assumption of one over the other (because we are exposing ourselves to equal risks no matter what position we adopt, and by taking a positive stance we might even be adopting an *increased* risk, as discussed above).

    scientific fine tuning

    I understand the concept you’re referring to here – and it’s entirely possible that you’ve used it consistently to mean “of all the possible universes allowed by theoretical physics, only a vanishingly small number allow [chemistry/life]” (I’m not trawling back to find an exception, should one exist – I’m happy to grant you the benefit of the doubt there!). I also completely understand the distinction between that and the teleological argument (which depends on the assumption that tuning of any sort is necessary), and I in fact deleted a paragraph on the matter from a previous post. A part of the point in objecting to that is the fact that “fine tuning”, scientific or not, is a loaded term which *implies* design. The concept of being “compatible with life” is a neutral term which describes our universe perfectly without any implicit assumptions about the origins of that compatibility.

    But my larger point remains that even this so-called “scientific fine tuning” is *not* a fact – it is a *hypothesis* which posits that universes which are *compatible with life* are extraordinarily rare, based on *assumptions* about physics beyond our own universe, and that “fine tuning for life” is a possible solution to that proposed statistical problem. I keep saying it, but *any* claim about *anything* outside our universe is a hypothesis which is, at present, unfalsifiable and such claims should therefore not be taken seriously. Again, you keep speaking of “possible universes”, when what you *should* be referring to are “proposed universes”. Just because a physicist can plug m(electron) = 5kg into his calculations and see what type of universe would result from that, doesn’t mean that a universe with a 5kg electron is *possible* – just like “Sally the married bachelor” is something I can propose but which is actually impossible.

    And since we can’t even get to (1) with any sort of justification, we can summarily reject the premises in the teleological argument which claim that “chance” and “physical necessity” aren’t the answer, because we cannot *at all* justify ruling those out when all we have are untested – and untestable – hypotheses about what it’s like outside our own universe.

    We test hypotheses using statistics

    What statistics do we have regarding the variability of universes? Given that we have access to precisely one universe, with no epistemologically valid justification for assuming that *even one* other universe does or could exist, how can we perform a statistical analysis regarding the frequency of different kinds of universes? And please, I’d ask you to offer *anything* other than “physicists have calculated that…”, because as I’ve tried to repeat time and again, the proposals of physicists (and the assumptions those proposals depend upon) are more math-y – but no more necessarily true – than the proposals of the Bible or any other holy book (or work of science fiction, for that matter).

    calculations have the shown following odds if the universe was generated randomly

    Again, “randomly” doesn’t necessarily mean “impossibly long odds”, unless you *assume* that all the constants are determined over a (near) infinite range of possible figures. As above – we can put m(electron) = 5kg into the formula; that doesn’t mean that such a universe is possible. Perhaps m(electron) *cannot* vary; perhaps it can vary over a very small range; or perhaps it can vary infinitely but *all* the constants are inextricably linked, such that if you know m(electron) you can calculate all the others in every possible universe. Heck, perhaps there are universes which don’t *have* electrons, but have entirely different physics that leads to lifeforms we could barely – or never – imagine (such as the disembodied intelligence – entirely unlike anything we have *ever* observed in our own universe – that many theists describe God as). The set of *proposable* universes, which comprise all those universes that we can suggest by assuming that physical constants can vary over a wide range, might be a much larger set than the set of *possible* universes that could or do exist given the laws applying to whatever reality might exist outside our own universe – to the extent that the set of *possible* universes, amongst all those that physicists have imagined, may even just have exactly one member.

    are you saying that most of modern Physics is wrong?

    Not in the slightest. I have the utmost respect for and interest in physics beyond cosmology and origins of the universe (though not enough time to delve as deeply as I might like) – and nothing but contempt for the many people who would do precisely what you’ve suggested and reject the lot despite its obvious successes. What I am, however, is prepared to take the “theoretical” part entirely literally (as theoretical physicists themselves do), and recognise that *theoretical physics*, whilst relying on math which does so well (yet amazingly imperfectly in many ways) in our own universe, is *theoretical* and is not epistemologically justified in making claims it cannot test about what goes on beyond its boundaries.

    That math can, depending on the assumptions you put in, give you a universe which comes about by physical necessity; it can give you a landscape where the overwhelming majority of universes are entirely incompatible with life; it can give a universe which “forks” at every event, such that all things that can happen will happen; it can give us a multiverse; it can give us a mathematical universe, where reality is an illusory consequence of a mathematical wavefunction (I don’t pretend to understand that idea); it can give us a universe which is nothing but a hologram (as is suddenly all over the news!). Given that the *same* math, with different assumptions, leads so many different places, why should we place confidence in *any* of the claims that are made?

    [a universe like ours is horrifyingly unlikely by chance] is a finding of modern cosmology

    The reason I asked this question goes back to something I noted in your previous post. You made several claims to the effect that “The possibility that [our universe arose] by chance is pretty low unless there’s a multiverse”. My question is, given that the *same* mathematics that posit the low probability of our own universe point equally strongly at the existence of a multiverse, how can you justify reliance on one of those propositions (low probability) whilst effectively dismissing – at the least, being extraordinarily cavalier about – the other (multiverse)? It seems to be an example of cherry picking, whereby you’re taking the bit that suits your hypothesis (low probability is *necessary* else God isn’t needed to deal with the teleological argument) whilst ignoring the part that would account for your concerns *without* your favoured conclusion.

    Let’s boil this down to a few dot point questions, see if we can tighten the discussion a little.

    1) On what basis can we justify the assumption that, outside our universe, physical constants are variable, or to what degree they can vary?
    2) Given that we have access to and knowledge about exactly one universe, and no access to or knowledge about the conditions outside our universe, how can we talk (intelligently) about the likelihood of other types of universes, or generate and analyse “statistics” about them?
    3) Given the myriad alternatives proposed by physicists on the basis of the different assumptions they can apply to the same basic mathematical theorems describing our universe, why should we assume *any* of them are valid or representative of reality?
    4) Given that you rely on cosmology in order to claim that our universe is unlikely, how do you justify dismissing the multiverse (or other proposals about the nature of reality outside our universe) that those same cosmologists propose? What basis do you have for rejecting the simpler multiverse (entirely physical processes) in favour of the far more extraordinary God (a supremely powerful and intelligent agent)?
    5) Why are you not concerned about the risk of an alternative hell, should your theology turn out to be incorrect? Why should the atheist have any concern for any of them, given that followers of any particular religion have no concerns about the hells proposed by any other religion?

    And finally, I really don’t think we can go further in this discussion until you understand [what cosmologists say about chance] and we can agree on it

    I completely agree that cosmologists make such claims. I understand their comments and the basis for them. I do think that a lot of them have been taken the wrong way, because they’re using the language of “fine tuning for life” to describe the fact of *compatibility with life* (which I think is more an unfortunate misuse of terms – like when a scientist lapses, in their professional speech, into common parlance and uses “theory” incorrectly – than evidence that they think the universe is *actually* fine tuned). I’ve read through most of the article you previously offered on the matter (I have a few pages left), and I was entirely unimpressed – rather than being the evidence for fine tuning, it was a large list of things we don’t know yet, and claims about the problems with various models and assumptions. There was no “evidence”, just criticism of modern physics and appeals to ignorance and the God of the Gaps. I think I did agree, though, with the criticism of some of Stenger’s arguments – but then again, so do lots of other philosophers, some prominent atheistic ones included!

    But I will absolutely accept the claims from cosmologists, regarding the improbability of the universe, the *moment* I have an epistemologically sound reason for believing that their claims represent empirically demonstrable facts about reality. I’m not trying to be arrogant or dismissive here – I’m trying to hammer home the point I’ve made, time and again, that what cosmologists propose is not the same as what is true, or even what is possible; and just because it can be proposed, doesn’t mean it should be believed. Going back again, *this is where the burden of proof lies* – without reasonable justification for *any* of the many claims made by theoretical physicists, or any reasonable way to test them all to distinguish between them, we shouldn’t *believe* any of them (or any model which is both less testable *and* less reliable within our own universe, either). We should withhold our conclusions until sufficient evidence is forthcoming.

    Furthermore, I would reiterate a certain hypocrisy in this – the cosmologists are the ones putting forth a multiverse (based on their *justified* and *tested* theories about our own universe); on what grounds can you justify ignoring their proposed multiverse in favour of “God did it”, whilst *relying* on the same physics to justify your claims about the imporbability of the universe, and simultaneously criticising *me* for disregarding physics (in which I, at least, am consistent in rejecting *all* of the untested claims [again, not claiming they’re *false*, but simply *withholding belief*] rather than cherry picking the ones that support my conclusion)?

    All the very best, as always, UnkleE.

  33. Hi Cheomit,

    I am going to stop commenting on fine-tuning. I’m sorry, but you keep saying things that aren’t scientifically true, or are not true of many of the facts I have offered. You keep saying things are speculative in various ways, and while there is significant uncertainty about some aspects, the majority is the subject of well known laws and equations verified by hard data and good enough to make successful predictions.

    I think you are going to have to read about it yourself. I suggest either Lewis and Barnes A Fortunate Universe or Martin Rees Just Six Numbers. Until then, there is little point in my continually saying “that is not what the science says” and you continuing on anyway.

    So I am going to go back to the point about risk and decision-making that you turned into a critique of Pascal’s Wager, and I think missed my point. And I want to do that by offering eight hypotheticals that deal with questions of uncertainty and risk, and asking you some questions, if that’s OK.

    (A) You are an evolutionary biologist trying to determine how a specific evolutionary process occurred (say abiogenesis or the formation of the eye). You have described a process that could feasibly have occurred, but have no way (at present) of demonstrating that it DID occur? Do you claim nevertheless that it did occur, or that we don’t know (the null hypothesis)?

    (B) You are an offical in the therapeutic goods administration charged with deciding if new drugs can be released onto the market. The drug thalidomide applies for release and you are unsure of its side effects. Do you say yes, no or require more data? Which is the null hypothesis?

    (C) Same again, but this time it is a drug to combat HIV/AIDS. Again you are unsure of the side effects, but those suffering from HIV/AIDS argue that they are dying and this is their only hope, so who cares about side effects? Do you approve release, or not or require more data? Which is the null hypothesis?

    (D) You usually walk home from work in an inner city office and take a shortcut down a narrow back lane in a poor part of town. Usually it is daylight, but one night you work back late and get to the top of the lane when it is quite dark. There is no obvious danger, but you feel uneasy. Do you take the short-cut anyway (the null hypothesis that there is no danger), or go the long way round on better lit streets?

    (E) You are a single hetero man who meets a beautiful girl who you relate to very well on a number of dates. She doesn’t believe in living together or sex before marriage. You have only known her a couple of months, but you feel that you love her. Do you ask her to marry you, or wait a little longer because you can’t be sure it will work out (the null hypothesis)? How much longer until you can disprove the null hypothesis?

    (F) Same situation, except the girl has hinted once or twice about a troubled past of sexual abuse and even a short stint working as a prostitute. You feel that you love her anyway. Do you walk away, ask her to marry you anyway, or wait longer (the null hypothesis)? How much longer until you can disprove the null hypothesis?

    (G) You are deciding about voting in a presidential election. You judge one candidate to be corrupt and self absorbed, but agree with some of his policies, while the other candidate has some good policies and experience but represents the political establishment which you don’t like. Do you vote for one of them, or a third candidate who you know has no chance of winning, or decide not to bother voting at all because you cannot know which course is best (the null hypothesis)?

    You are welcome to give me your answers, but my main point is to ask these questions:

    1. A while back you offered the principle: ”For practical purposes, of course, we often need to act in the absence of epistemologically valid justification – but in those cases, we *always* default to the null hypothesis”. Would you default to the null hypothesis in all of these cases?

    2. If not, which ones would you do that for?

    3. What other principle would you use in any other cases?


  34. Hi UnkleE,

    I’m sorry, but I take exception to your claim that I “keep saying things that aren’t scientifically true, or are not true of many of the facts [you] have offered”. Firstly, I’m not defending a scientific position, I’m defending an epistemological one – never adopt a belief for which there is insufficient evidence. Certainly, there is science at play here informing the discussion, but science only goes so far and it doesn’t get us anywhere near where your argument needs it to. Science, at present, has no way to address supernatural claims or to investigate claims about what is or is not possible outside our universe. This is *not* controversial, yet fine tuning arguments require it to be false.

    As I noted previously, we have *two* relevant facts about the universe – it exists, and it is compatible with life. To get to the teleological argument, you need to demonstrate a *need* for fine tuning – that a universe which is compatible with life is surprising (i.e. not physically necessary) and that chance is impossible. But all you have as justification for that conclusion is incredulity and the predictions – which is all they are – of theoretical physics. But that’s not good enough – the vast majority of all predictions made by scientists of all stripes have been shown to be false, and most predictions made by theoretical physicists today are mutually contradictory. Until we have empirical data which shows us that the physical constants could be other than what they are, or that some other universe (or universes) actually exist, I’m sorry, but you have *no* justification for the claim that the universe is surprising, and thus none for the claim that it must have been designed. Those are claims that the theist (or the physicist) must establish, and whilst I’m not deep in the field, I feel quite confident that I would have come across an experiment which proved the existence of another universe, that the speed of light in a vacuum is variable, or some other such finding.

    Consider also the Higgs boson. Ten years ago, the Higgs *existed* as a part of reality. We had solid math that predicted its existence. We had a coherent model of physics that described our universe well, but depended upon the Higgs. Physicists were confident that they would find it. But we were not *justified in believing that it existed*, because all we had were predictions, and predictions are not evidence. Most predictions turn out to be false, after all, and empirical evidence is the only thing we can use to figure out which ones are reliable and which ones are nonsense. Indeed, there were multiple competing predictions about the precise nature of the Higgs, and most of them were rejected after the evidence came in. This whole argument about fine tuning is no different. All we have are predictions and models, with no empirical evidence that the constants can vary, that other universes are possible, that our universe is at all surprising to anyone who knows enough about physics. Until someone comes up with and performs an experiment that demonstrates any of those things, the fine tuning argument depends on nothing more than assumptions arising from assumptions and appeals to ignorance, incredulity and the god of the gaps (as repeated over and over in the Barnes paper you previously cited).

    Let me offer one more analogy. I gave my son ice cream one day – plain vanilla. He loved it, of course. He asked for ice cream again and again and again. To him, at that time, ice cream meant *vanilla* ice cream, and it would be absurd for him to have been surprised to find vanilla ice cream in his bowl of ice cream. But one day, I put chocolate ice cream in there. He was overwhelmed with surprise and excitement – he never even considered that chocolate ice cream was possible! We’re not talking about anything different when it comes to universes – we have a big tub of vanilla universe, and we’ve never seen any other flavour in our bowl. We don’t get to be surprised by that, though, because every time we’ve gone looking, all we’ve found is vanilla. Unlike my son, we’ve dreamed of other chocolatey universes – but we’ve never actually found any chocolate, or even any evidence that chocolate is actually out there somewhere! If we find a big old scoop of chocolate ice cream one day, that’ll be *awesome* – but I’m not going to pretend that it’s out there just because it would be delicious.

    You also failed to explain how your reliance on the untested predictions of theoretical physics regarding variability of physical constants, whilst casually dismissing the equally well supported predictions of a multiverse (which eliminates the need for fine tuning), is not cherry picking (or how it is scientifically or philosophically more justifiable than my withholding belief in the absence of empirical evidence to support one of the many competing hypotheses).

    But given that the fine tuning argument was entirely a red herring to the original topic, I’m largely fine with moving on.

    The examples you’ve offered are largely uninteresting. To all of them, I would state that we reject the null hypothesis for *practical* purposes when the evidence is sufficient to convince a reasonable person that the claims are true. “Reasonable person” might vary from one scenario to the next, and would often be a person appropriately versed in the relevant art, but that’s a pretty basic standard to aspire to. I’m not going to go through them individually – not least because you’ve done a poor job of identifying null hypotheses (for instance, the drug regulation examples require us to establish *at least two* null hypotheses – “this drug will not provide benefit” and “this drug will not cause harm” – and balance the evidence and expectation of each against the other).

    But, for all cases, we reject the null hypothesis for practical purposes when the evidence supports us doing so. Without evidence, I’d remain neutral in all of these examples – certainly, as written, A demands we remain neutral (because we’ve got naught but a hypothesis); B, C, E and F are complex multi-part questions which need far more info that you’re offering; D I keep walking. G is a poorly framed hypothetical – are you asking me to reject the null hypothesis “I do not prefer any candidate”, which is easy and relies entirely on facts (mental states) about which we have indisputable knowledge; or are you asking about deciding which candidate will have the best outcomes – which requires the consideration of *multiple* null hypotheses (“Fred will not be the best candidate”, “Sam will not be the best candidate”, …) and a weighing of the evidence for each?

    But here’s the key point – if we assume you mean, in G, “I do not prefer any candidate”, then it is only in that scenario that we can *actually* reject the null hypothesis on epistemological grounds, because only in G can we *know* the truth (preferring one candidate over another is a mental state, and we are absolute authorities on our own mental states). For *all* of the others, you’re relying on imperfect information in order to guide action, and whilst that’s necessary for life, it doesn’t mean that we can claim knowledge about the situations (you walk down the alley and don’t get mugged – because there was no mugger, or because the mugger decided you weren’t a good mark?). Thalidomide also provides a perfect example of the principle of updating your practices on the basis of new evidence – having been removed then brought back to market with new restrictions – which is hard to do if you adopt epistemological certainty about an issue.

    Even the empirical claims about which we are *most* certain are only highly probable. When CERN announced the discovery of the Higgs, they noted that they have “five sigma” confidence. That means that the probability that their experiments would have given the results they did *if the null hypothesis is true* is one in 3.5 million. Note carefully, however, that they’re not saying it *does* exist – they’re saying that the probability that the data they obtained would have happened even if it *doesn’t* is very small, and thus they’re quite *confident* that it does exist. Effectively, the statistical analysis *assumes that the null hypothesis is true* and figures out what the chances are that they got a false positive. So when you look at rigorous empirical science, you should note that *nobody* rejects the null hypothesis fully – they cast their results in terms of the null hypothesis actually being true!

    None of these examples, however, are remotely analogous to the question of hell, and whether we should modify our beliefs about god in light of it. Firstly, the examples you’ve offered are all real world issues, and we can be reasonably confident about the existence of certain consequences (death, humiliation, personal injury, a broken heart, Donald Trump). But when it comes to hell, we have nothing at all comparable to that! We do not *know* that hell exists, let alone the criteria you need to meet in order to end up there! We have no evidence of any sort of “life after death”, let alone of a place your hypothetical “soul” ends up! We have nothing more than *stories* – and multiple contradictory and mutually exclusive stories at that. Where are the empirical facts about reality which demonstrate that one story is more reliable than any of the others?

    In turn to you, I would reiterate the question that you’ve dodged twice now: do you adjust your behaviour to the proposed risk of an Islamic hell? If not, why not? Given that you’re engaged in worship of a false idol, what reason do you have to believe that, if we’re both wrong, Allah would treat me more harshly than he would treat you?

    I’d also offer my own analogies. 1) You’re standing on a small median in the middle of a busy road. A single step will put you into traffic – and could result in anything from a minor bruise to death. From the side of the road, a billion bystanders yell to warn you that there’s an invisible, silent, massless dragon hurtling toward you from above, intent on eating you alive and tormenting your soul for eternity. They tell you that if you move it will crash into the ground and be unable to get you. At what point do you take a step?

    2) You’re deprived of your senses and standing on a tiny precipice above an enormous canyon. You *know* that there is *one* tiny pathway to safety – but if you miss it, you fall to your death. Which way do you step?

    3) You’re on a precipice, deprived of your senses. You do not know if there is a pathway out at all (perhaps you’re atop a tall, narrow skyscraper). You do, however, have everything you need to live a fulfilling life without ever needing to move. Do you step? If so, which way?

    So, to be entirely clear, my questions to you are:

    1) Point me to empirical evidence (not predictions or gaps in our knowledge) that demonstrates our universe is in fact surprising. I want experiments or data which show that *some particular set of predictions* are worth taking more seriously than all the other sets out there.
    2) In the absence of that empirical confirmation, how do you justify relying on predictions that the universe is surprising whilst rejecting equally well supported predictions that it isn’t?
    3) Do you adjust your beliefs in order to account for the risks posed by alternative hells? If not, why not?
    4) And, of course, responses to my own scenarios.

    All the best, UnkleE.

  35. Hi Cheomit, I’m going to focus on your answers to my questions. From your comments, I get the following:

    1. A while back you offered the principle: ”For practical purposes, of course, we often need to act in the absence of epistemologically valid justification – but in those cases, we *always* default to the null hypothesis”. Would you default to the null hypothesis in all of these cases?
    2. If not, which ones would you do that for?

    You say the principle is: “for all cases, we reject the null hypothesis for practical purposes when the evidence supports us doing so.” You then go on to say you wouldn’t draw a conclusion for A. You say there isn’t enough information in B & C, but this is a hypothetical and we assume you know this information. So I’m guessing this means you would want more information for B. But what about C?

    You are drawing a definite conclusion with D. For E & F, of course you lack all the information, that is the point, but you seem to be saying that you wouldn’t make a decision to marry without more information, which might mean you’d never marry.

    I’m not sure I can see any answer to G.

    So it seems that you would take action beyond the evidence with D, it is unclear with C & G, and the rest you’d do nothing without more information, even if it means you’d never marry.

    3. What other principle would you use in any other cases?

    So I’m interested in what principle you use for D? (And also how you’d decide C & G.)

    “None of these examples, however, are remotely analogous to the question of hell”

    You seem to have a strong interest in hell. But I never mentioned it, and I don’t believe in a hell as you mean the word. It is irrelevant to what I am saying here.

    In answer to your hypotheticals, my answer is “I don’t know”. I’d try to work things out and then decide what I thought was safest. But posing these hypotheticals suggests to me you’ve missed the point of mine. I wasn’t trying to set up some sort of difficult scenario like a trolley problem. You had made some definite statements about your epistemology, so I wanted to test how you dealt with uncertainty and risk. But I haven’t made any strong statements beyond suggesting that uncertainty and risk should be considered. Your hypotheticals illustrate that.

  36. Hi UnkleE,

    1) You don’t really seem to be understanding my stance here. You need to separate epistemological certainty (which should be *rare*) from practical considerations. On a practical level, you almost never *disprove* the null hypothesis completely – you gather evidence to the point that the null hypothesis is sufficiently unlikely (for various levels of “sufficiently”, depending on the question), and you act in accordance with that, but you don’t get to presume that the null hypothesis is actually false. As I pointed out above, scientific results are presented as presuming that the null hypothesis is in fact true, and whilst lazy common language distracts from that, it doesn’t change the fact at all that science never truly rejects null hypotheses (it just gives us some degree of confidence that the null hypothesis is unlikely to be true given the evidence we have).

    So, for marriage, a scientist should assume that the girl doesn’t love him until he has enough evidence to justify the conclusion that she does, or at least until the evidence that she does sufficiently outweighs any evidence that she doesn’t, or the benefits of marrying outweigh the risks of the relationship failing (there are multiple ways you might want to test something like this). But a simple example – I *do not believe* with anything like absolute certainty that my wife loves me. I am *very confident* that she does, because I have substantial evidence to justify that conclusion, but I know it’s entirely possible that she’s lying to me. The benefits of acting in accordance with the assumption that she loves me also outweigh the risks of loving her and not being loved in return, though, so that’s how I act. It’s a very simple principle. I act *against* the null hypothesis – whilst not rejecting it fully – because the evidence justifies my doing so (it would be far more surprising to a reasonable outsider that she didn’t love me than that she did).

    For D I am in fact acting in accordance with the null hypothesis (there is no heightened risk) in the absence of evidence that there is. I’m not at all certain that there’s no risk, but I have no reason to believe that there is and I’m not going to change my behaviour simply because I have a vague feeling (a reasonable outsider would not conclude that my feeling of unease is a good reason to abandon the null hypothesis and justify taking the long way).

    For A, all you offered is a hypothesis – and unless you can test that hypothesis (which may not be possible – often isn’t for historical evolutionary propositions of the type you’ve given), you can’t establish it has any basis in fact and should therefore not believe that it did in fact happen that way (a reasonable outsider would not have any reason to believe that you’re doing anything more than telling “just so” stories unless you can test them).

    For the drug examples – you’re stepping deep into my actual field. But really simply – when the evidence of benefit outweighs the evidence of harm, we license the drug for sale. We then monitor it in the patient population to establish any previously undetected harms and adjust the licensing as required. Whilst the FDA, for example, are human and sometimes bow to public or political pressure, they generally do not care what patients *want* or *think* works – they rely on objective data in their decision making, not personal opinions and anecdotes. Thalidomide is a great example – the evidence justified its licensing (except in the US, although that may have been the luck of having a conservative reviewer rather than good management); the evidence of birth defects caused a revision and removal from the market; and further studies have shown it to be useful for treating certain infectious diseases, leading to it being re-licensed (with a restriction against use in pregnant women). The HIV drug would be no different to that at all. We license drugs for use when a reasonable outsider would conclude that the benefits outweigh the risks – and we revise those decisions in the light of new evidence.

    For G, I again put it to you that your hypothetical is unclear. If all you’re asking is “how do you decide which candidate you prefer?”, that’s easy – I can be epistemologically certain about my own mental states; preferring one thing over another is a mental state; therefore I can be epistemologically certain about which I prefer (or if I prefer none of them). If you’re asking the question about how you justify reaching a preference in the first place, though, it’s *no different* to the other examples – just more complex. You’re required to interrogate reality in an attempt to discover which option leads to the best outcome, and that involves a lot of risk-benefit comparisons between candidates, desired outcomes, personal values, and so on – but it’s still hypothesis testing (“will X or Y offer the best outcomes for the economy?” is decided by, say, comparing the results of the thought experiments “X will lead to better outcomes” versus “Y will lead to better outcomes”). You (provisionally) reject multiple null hypotheses on a probabilistic basis (using evidence gained from candidate statements, thought experiments, personal values, candidate history, and so on) until you reach a preference – or determine that you cannot, of course. One last time, you cast your vote on the basis of a complex synthesis of evidence that would lead a reasonable person to conclude that voting for candidate X is most likely to lead to the outcomes you favour.

    I actually kind of like G, though, as an analogy to the topic of god. I can ask, “do I believe god exists” (“is candidate X the best choice”) – and I can answer with certainty “no”, because belief is the mental state of accepting that a claim (“god exists”) is true (and I am therefore an atheist – one who does not believe in a god). I can ask “does god exist” (“is candidate X better than candidate Y”), and my answer is “yes or no, but I don’t know which” (making me an agnostic as well). I can also ask “does the evidence suggest that *does* god exist” (“does the evidence suggest that candidate X is better than candidate Y”) – I obviously think not, for all definitions of god; and “does the evidence suggest that god does not exist” (“does the evidence suggest that candidate X is *not* better than candidate Y”) – I’ve previously stated that I think yes to some degree, for certain definitions of god. The whole topic of god’s existence is confused when people – theists and atheists – conflate these questions but focus on the second one (“does god exist” – and, mostly, “does *my* god exist”), and ignore the nuanced differences between (in the order I presented them) belief, fact, and reaching conclusions on the basis of evidence (evidence itself being a whole other thing that is constantly misunderstood).

    As to hell – consider it both a shorthand for “what happens after death” (whatever you happen to believe that is), as well as the most extreme example of a hazard that I think *anyone* can propose (is there a greater hazard than eternal torment?). If you’re offering up the principle of assessing risk in order to modify behaviour, hell is the strongest example you could possibly appeal to, and so my use of a literal hell in my comments is a form of steel-manning (as opposed to straw manning) your arguments.

    Your concern here also throws another of my previous questions into a rather bright spotlight – you’ve proposed that risk should modify behaviour (or, very importantly, belief), and yet you reject the notion of the *greatest conceivable hazard* – hell. How do you justify that stance in light of your risk management principle? Shouldn’t you act – and modify your beliefs – in order to account for (mitigate the risk of) the greatest conceivable hazard? If not, why not? And if so, how do you decide which of these competing “greatest conceivable hazards” you should focus on avoiding?

    But as to your response to my scenarios, “I’d try to work things out and then decide what I thought was safest” is a perfectly reasonable practical principle (not entirely dissimilar to what I’ve advocated), but my concern is with *how* you make that assessment in the cases I outlined – that’s what you need to defend. At what point do you decide that it’s safer to step into traffic than risk the dragon? How do you justify your decision on a *practical* level, let alone an epistemological one? At what point do you decide that turning 23.64° to your left and taking a step is the safest path off the first precipice? At what point do you decide that there *is* a path off the second, let alone where it lies? I *really* want an answer to those questions, because they go to the very heart of your principle.

    Thanks again for the continued interesting discussion, UnkleE.

  37. Hello again. I don’t want to get into a “Yes I do”, “No you don’t” pattern but I think I do understand you when you say: “You need to separate epistemological certainty (which should be *rare*) from practical considerations.”

    And I think the same. We can never be certain of anything because we cannot prove we are not dreaming, or brains in a vat, or that the external world is real, or that our cognitive faculties are reliable, etc. But leaving aside those sophistries, we can only “prove” mathematical and formal logic statements. Everything else is known with a degree of confidence or probability only.

    It is because I do understand that I ask you to explain how you make decisions in the face of uncertainty, which implies some risk of being wrong.

    I’m sorry I misunderstood your answer to D, but no matter. I think these are what I was looking for, statements of how in principle you would decide in the face of uncertainty: “You need to separate epistemological certainty (which should be *rare*) from practical considerations …. I act *against* the null hypothesis – whilst not rejecting it fully – because the evidence justifies my doing so .” You also talk about the benefits outweigh the risks.

    And again I agree. I think that is how I live my life. And that is how I develop or revise and/or maintain my belief in Jesus. I don’t think it is certain that Jesus is truly divine, but I live according to that belief and therefore against the null hypothesis of unbelief because I believe the evidence justifies that choice, especially when I consider the benefits and the risks.

    You have criticised some aspects of what I have said, especially for, allegedly, flirting with Pascal’s wager, but you have offered the same principle, just your version of it doesn’t have the scare words of “Pascal’s Wager” and “hell”! 🙂

    So I would respond to your hypotheticals in exactly the same way as I think you would. I would assess the facts and which ever choice leads to the benefits outweighing the risks, as much as I could assess them, I’d make that choice.

    Of course in reality I’d probably be panicking and running on adrenaline, and that illustrates another matter I haven’t raised yet. We like to think we decide things rationally, but often we don’t, often our hormones or nerves or whatever take over.

    I’d be interested to know a little more about the person I have been discussing with. I gather you work in medicine or organic chemistry, and I presume your comments about marriage were real life and not hypothetical, but I have no idea of where you live, what else you think, how you got to be where you are today as an atheist/agnostic, etc, if you’re interested in sharing a little.

    I am 71, a retired civil engineer and environmental manager, married with 3 adult children, I live in Sydney Australia, and while I wasn’t brought up christian, I have been a believer since my mid to late teens, about the time I began university. I enjoy the web, some sport, alternative pop music, reading (mostly non fiction on topics I blog about), travel and meeting new people. So that’s me.

  38. Hi UnkleE,

    Thanks for clarifying. It did seem as if you were conflating the issues of epistemological certainty and empirically justifiable confidence, but perhaps I was misreading you. I was going to point to the problem of hard solipsism as a perfect example of that in my previous response, but decided not to – glad to see you recognise it anyway!

    Let’s also be clear here – we need to separate emotional responses from rational ones, because the former aren’t particularly useful when we’re looking for external, objective truth. What one feels is a perfect guide to *what one is feeling*, but it’s a horrifyingly flawed way to get to the truth about much else. As far as this question is concerned, we have plenty of time to calmly consider it, so we *should* approach it rationally and not let emotions get in the way.

    As to me – I have a PhD in organic chemistry, and whilst I still work as a chemist, I’ve moved to a rather different area. (Funny story, though – the user name was created with a random generator, rolling until something that “looked like a word” came out. It was only later that I recognised the resemblance!). I have a number of other degrees, in biomedical sciences and philosophy. Married with a child. My family is nominally Catholic (but I only really discovered that rather recently – religion never played a part in my life), and I’ve always been an atheist in retrospect – although I only owned that relatively recently, and went quite some time considering myself agnostic (thinking that was the intellectually honest label – which it isn’t, of course, it’s the cop-out that people use because “atheist” has been so demonised).

    So, back to the argument. I’d have been entirely surprised to discover you live your life in any way different to the type of risk-benefit method I put forth – that’s how most people (try to) make a significant proportion of their carefully considered decisions (with a good heaping of irrational, emotionally-driven choices alongside, of course!). Your principle is similar to mine, although I think it might put a greater emphasis on risk than mine necessarily does (of course, there are times where we absolutely do modify for risk – drugs being again a neat example, where we know *a lot* of the things that can go wrong, and so we specifically test for them even if we have reason to think they’re unlikely to be a problem with “drug X”).

    But I cannot see how you can apply that principle to a supernatural hazard like hell or the afterlife or whatever. Both our principles are entirely applicable to empirically testable situations, where we know, can measure or can reasonably anticipate risks and probabilities based on reasonable analogies, but they fail entirely when we get to epistemological justification for beliefs and for changing our actions in light of hazards for which we have no analogy, for which we have no reliable evidence, and for which there are multiple equal, competing and mutually exclusive claims. As I said, hell is the *greatest imaginable hazard* – and the existence of such a hazard would warrant fairly extensive efforts at mitigating the risk. But the whole point is that there is no reason to think any such hazard exists – and without a hazard, there is no risk, and any efforts to mitigate a non-existent risk are all costs and unnecessary hazards themselves!

    The point of my scenarios was to tease that out. To be clear – I’m going to use the “royal you” here – please don’t think that I’m accusing you of anything I’m about to put forth!

    If we consider the first – the dragon, obviously, is our vengeful god. We have no reason to think that dragons exist – we’ve never found anything even reasonably close to them in the real world, after all (it’s not like the black swan problem, where we had swans but not black ones – if it were that simple, we’d simply point to dinosaurs. But dragons are magical creatures that breathe fire and have complex language and hoard treasure in massive caverns, and no creature we’ve ever found is anything like that). We’ve no evidence of (large) invisible things being possible, let alone of living things that are massless (all our evidence points *against* those things)! Then, we get to the practical considerations. We have no reason to believe that dragons do eat people, or that they can capture your soul (that your soul exists…). The *only* evidence we have is all those bystanders. But people are easily fooled, as any magician can show you, and if they’re shouting about something that is, to the best of our knowledge, impossible (magical and massless being enough to justify that label for empirical purposes), you’re probably not justified in trusting them. God is no different – something not only entirely different to anything we’ve ever been able to find or reasonably imagine might be true, but something that is by definition beyond anything with which we can ever be familiar and which is often defined as being entirely beyond our comprehension or imagination. God is a *huge* extrapolation from even my dragon, let alone anything we’re actually familiar with!

    On the flip side, of course, is the very real risk of the road. We know what happens when you step into traffic – and whilst we don’t know how severe the consequences will be in each individual case, we can predict how bad you’re going to get hurt. When it comes to religion, your costs might be time or money – a minor bruise – or they might be your life or the lives of others (a monastery, a suicide vest, or kicking your gay son out of the house to die homeless under a bridge). Even if you escape with a bruise, you might cause someone else harm as they crash into a wall after swerving to miss you (your local church gives your tithes to the higher churches, who in turn fund terrorists or use it to protect pederasts; or widespread support for moderate religion provides the extremists with a shield against criticism). So, the costs of mitigating the risk of hell are *not* zero, and can be severe. If hell doesn’t exist, you’re hurting yourself or others for no possible benefit – an outcome which is ethically abhorrent by any standard I’ve ever come across.

    The second and third analogies get at the competing hazards out there. In the second scenario, I’ve granted that there’s a way out – that one of the myriad gods that people have imagined *does* exist and if you believe right you will be saved. But in the absence of any reliable way to determine *which* is the real god, how can you justify choosing one over all the millions of others? Realistically, it’s not significantly different to the problem you repeatedly cite with the universe’s being compatible with life – assuming I grant your hypothesis that there are 10^10^123 possible universes, the chance of rolling ours is miniscule but non-zero; likewise, your chance of “rolling” the right god is better than that, but still bugger all! And let’s be blunt, most people don’t even *try* to find the “real” god – after all, a person’s god of choice is most tightly correlated with where they were born, and even conversions from one sect within a broader religious category (let alone from one religion to another) are rather rare.

    The third situation, of course, represents our reality. Anybody who claims to *know* that a god (let alone their god) exists (scenario 2) is simply wrong. One might exist, but we can’t know either way (hard solipsism proves that), so we are absolutely living in scenario 3. And if we couldn’t reasonably decide where to step in scenario 2 (where we knew there was a *chance*), how can we possibly do so in 3? Surely the only thing you can do is live your life as best you can, with an open mind which is prepared to step as soon as you can justify doing so – but not a moment before, else we’re taking an existential risk for no reasonably likely benefit.

    I couldn’t come up with an analogy for competing consequences, unfortunately – e.g. you can choose between Christian hell, Islamic hell, and neutrality (atheism) where it turns out that one of the gods exists (say, Allah so the atheist and the Christian are wrong), to examine why the atheist position remains rational (i.e. why the atheist would be in no worse a position than the Christian – and perhaps a better one).

    So, I think we may agree that empirical principles cannot provide epistemological justification for belief (confidence, yes; belief, no). I’d still be curious to know how you’re applying your principle to whatever afterlife it is you think is worth being concerned about, instead of adjusting your actions to account for hell – the greatest imaginable hazard. I’d also like to know how your principle can apply when we have all those competing hells (and less severe or even more beneficial afterlives) to choose between.

    I’d also be curious to know whether you think your belief in god is *actually* influenced by the risk management principle you described.

    All the best, UnkleE.

  39. Hello again, thanks for the personal information, it helps me understand where you are coming from.

    “But I cannot see how you can apply that principle to a supernatural hazard like hell or the afterlife or whatever. “

    I think our choices are the complex results of our “priors” (our genetics, upbringing, presuppositions, etc) and factors such as evidence, logic, emotion, risk-aversion, prevailing views on related matters, previous decisions on the matter, etc. I think any simple explanation is suspect and generally mistaken, though of course we have to simplify a little in discussions like this. I think strong claims to evidence and rationality (like I felt you were making at the start of this discussion) don’t account for what psychologists tell us about our decision processes. So let’s keep all that as a proviso in what else I say.

    Risk = consequence x likelihood. There are many, many enormous hazards that could possibly occur ( the sun stops shining, the earth falls into the sun, nuclear war, getting run over by a truck, getting cancer, going to hell, etc) and if we worried about them all, we would totally spoil life. So it makes sense to eliminate as many as we can. We can do that by considering likelihood, and also risk of alternative actions. Without going into every detail, I think most of those hazards are unlikely, or there is little I can do about them, so I press on with life. Everyone does this to some degree or other.

    My original point was simply that there are hazards associated with belief and disbelief, the likelihoods are not vanishingly close to zero (I think, and I think you think so too from what you have said), and there are ready alternatives. To be more specific, belief in God can have some clear advantages in this life and the next, we cannot say the probability of God (likelihood) is zero, and it isn’t too difficult to check things out with a slightly more holistic set of criteria than empiricism. This doesn’t mean we can and should manufacture belief, but it does suggest to me that it is sensible to give God a little bit more thought than many atheists/agnostics do.

    “I’d also be curious to know whether you think your belief in god is *actually* influenced by the risk management principle you described.”

    Psychologist Jonathan Haidt says that we make our decisions on religion, ethics, etc, intuitively and then rationalise them later. I don’t know if I go along with him completely, but I’m sure there is truth in that. Also, I came to my beliefs over a period of time and have since kept refining them. The reason why I originally believed may be different to why I now believe.

    But having said that, I think so. Except I don’t think of it as “risk management”, but rather as uncertainty. So my beliefs very much are built on how to decide in a situation of uncertainty. There are many things in christian belief I am uncertain about, some things I used to be uncertain about that I have since resolved, and vice versa. In situations of uncertainty, we either take action (which requires some sort of leap, of faith, or hope, or probability, or whatever we call it) or we remain paralysed. I choose to choose and act rather than remain on the fence over many things (generally the most important things) while remaining open-minded about others.

  40. Hi UnkleE,

    I largely agree with your assessment of how people come to their beliefs – we do, in many cases, retroactively develop a justification for an already held belief, rather than developing a belief on the basis of evidence and reason. That said, we *can* form rationally-based beliefs if we take the time to do so – and we should, wherever possible, particularly for such important questions as the existence (or not) of a god. I also agree with your description of how we tend to deal with extreme existential hazards over which we may have little or no control. This, of course, might not always be rational – after all, if we lobby hard enough, we can convince our governments to invest in the technologies required to, say, intercept or destroy and Earth-bound asteroid – but it’s understandable, and I think you’ve more or less captured what we generally do!

    As to belief or disbelief in the supernatural, however, I continue to disagree with your analysis. There is a hazard associated with non-belief in god/afterlife/whatever *if and only if* hell (or some other consequence) exists – if there is no such consequence, then there is no risk. Now, obviously we can stick with hell and say that the *hazard* would be infinite (and thus any non-zero likelihood would result in infinite risk), but if there is zero likelihood of the hazard occurring (because it doesn’t exist), then there is no risk to be accounted for. In contrast, we have very well known hazards and likelihoods associated with belief – as I said, ranging from the financial costs of tithing or donating to less effective religious charities over more effective secular ones, all the way to existential hazards to ourselves or others (monasteries, self-harm, suicide attacks). You might argue that taking a gamble on that infinite risk is worthwhile, given the palty (in comparison) consequences of belief – but that’s not a rational decision. And just because people make irrational decisions all the time, doesn’t mean that we should do so deliberately!

    On the flip side, of course, you point to the benefits of belief. In the case of benefits in the next life, of course I’m going to disagree with you – you have no evidence that there *is* a next life of any sort, let alone one in which your actions here and now will determine your status therein. But I also disagree with your assertion that religious belief offers any particular benefits in *this* life that cannot be obtained through non-religious pathways. None of the benefits associated with religious observance seem to exceed the benefits available to members of other similar types of non-religious communities – that is, it is the *community and social* aspects of religious observance or other groups that provide benefits to individuals, rather than the religious observance itself. This is borne out by noting that being more religious does not correlate with improved wellbeing relative to the less devoted members of your congregation – and I could be wrong, but I recall that the correlation may in fact be inverse (more devout = less wellbeing). I’ve seen your list of medical studies into the effects of prayer – but there’s not a single study out there that provides even remotely convincing evidence of medical benefit (and I also note that for one of the more amusing results, the STEP trial by Benson et al., you only cite *half* the conclusion in order to list it as a study showing “no significant response to prayer”, when in fact it showed that there were significantly *more* complications in the group which knew they were being prayed for! Of course I understand medical research better than to pretend that this result will likely stand up to further testing, but it is very interesting that a Templeton-funded study gave evidence of harm, and that you’ve chosen to blatantly cherry pick from the results in order to present the best possible picture for your position).

    So, even on a practical level, I cannot see how risk can be used to justify belief in the supernatural. There is no sufficient reason to believe that a hazard exists; there is solid evidence of minor to existential risks associated with belief; there is no evidence that there is a “next life”, let alone that we can do things to benefit ourselves therein; and there is poor to no evidence of specific benefits in this life (and definitely no good reason to conclude that any such specific benefits outweigh the known harms). Given all that information, the rational analysis would surely be to abstain from belief, to maximise known benefits and minimise known risks, given the absence of evidence for the supernatural risks and benefits you propose.

    And *all* of that still fails to acknowledge the fact that, when it comes to supernatural hazards, multiple competing and mutually exclusive ones have been proposed – yet you’ve chosen to focus in on *one* of them. You have acknowledged that you believe in some sort of afterlife with consequences for non-belief, but not in a hell (the greatest imaginable hazard, and thus the one you should seek to avoid if consequences in the next life play any part in your reasons for belief – which, again, you have claimed that they do). Many Christians *do* believe in some form of hell – be it the lake of fire put forth in Revelation, or the notion of hell as being separated from God’s grace, or whatever. And Muslims, Buddhists, Hindus and so on all believe in different afterlives with different consequences and different rules for what you need do to avoid them. Given all of that – and given that there is clearly no evidence or argument that can be made that is persuasive to a reasonable person who does not already subscribe to a reasonably similar set of beliefs (otherwise almost everyone who’s ever considered the matter would be, say, a Sunni Muslim) – how can we possibly make a reasonable choice between them, even if we decide that one must be real (which, again, I obviously do not agree with)?

    This is why your principle of belief in order to mitigate risks fails. You cannot establish that any hazard (and therefore a risk) exists; you cannot distinguish between the various hazards that have been proposed; you cannot be at all confident that your beliefs don’t actually make your situation worse than mine is; and you fail to apply your principle in a rational fashion by acting in accordance with the greatest of the proposed hazards, favouring instead to focus on a substantially lesser one. Your principle, in a case like this, results in adopting known risks and forgoing known benefits without any rational reason to believe that you’re mitigating any hazards or gaining benefits. The idea of modifying behaviour in light of risks is perfectly valid when applied to known or reasonably anticipatable hazards (like the ones you presented in your eight scenarios a few posts back), but it is not a useful method of responding to multiple competing unknown and not reasonably anticipatable hazards.

    Finally, I’d turn back to the epistemological point – modifying your behaviour to account for a risk is one thing, but modifying your beliefs is something else entirely. We do not, after all, choose our beliefs – we are either convinced that a proposition is true or we are not, and that’s an automatic process. We can sometimes stack the decks in order to try and obtain a specific belief – by seeking out valid evidence, for example, but we can’t actually *choose* to believe something – at best we can pretend to do so. This fact is why I asked about whether or not you personally think that your “uncertainty” actually contributes to your beliefs – I’m happy to take you at your word that you think it does, but that would make you rather unusual (most people I’ve heard answer that do so with something to the effect of “well, no, but…”). It seems most people offer this as a reason for *non-believers* to consider, without it being a reason for *their* belief – and it *cannot* be a valid reason for belief (as opposed to *behaving as if you believe*), because we cannot be afraid of something which we do not believe exists. From an epistemological point of view, this argument is necessarily circular. Why should we believe? To avoid eternal consequences. What are those consequences? Hell (or whatever). Why should we believe in hell? Because it’s the only way to avoid it. You cannot tell me there’s a rational person alive who would see it in any other way, and who would be convinced by the appeal to fear of hell if they didn’t already believe that hell existed.

    I’d also argue that I am in no way paralysed by my lack of belief – the only thing I, as an atheist, cannot do is provide a firm answer to the question “does God exist?”. I can answer *every* other relevant question carefully and thoughtfully, I can consider new evidence should it be forthcoming, I can change my mind either way if it is justified by that evidence, and I can live my life in an ethically sound and intellectually honest way such that any god who would condemn me for exercising the rational skills he granted me wasn’t worth my time in the first place (being made as I was, I could never have lived up to his requirements to believe in the absence of evidence which would convince a reasonable outsider). I have to ask, do you genuinely think you can do the same? I have already given some examples of things which would make me shift my views (if not express complete confidence in the existence of god – but then, I don’t have complete confidence in the existence of the chair I’m sitting in!); but let me ask, do you know what would make *you* change your mind and give up your beliefs? If you don’t, doesn’t that suggest that you are in fact more paralysed by your belief than I am by my lack thereof?

    Anyway, I’ll stop there. Apologies for the again slow response – it has been a busy week! All the best.

  41. Hello again, don’t worry about delays, it’s best to be free to communicate as we each have time.

    ”There is a hazard associated with non-belief in god/afterlife/whatever *if and only if* hell (or some other consequence) exists – if there is no such consequence, then there is no risk. Now, obviously we can stick with hell and say that the *hazard* would be infinite (and thus any non-zero likelihood would result in infinite risk), but if there is zero likelihood of the hazard occurring (because it doesn’t exist), then there is no risk to be accounted for.”

    Your analysis here makes a couple of assumptions I question.

    1. You seem to always bring this question back to hell (which I haven’t been mentioning) and then assuming it is an infinite hazard. But there are many other possibilities – e.g hell could be finite, or it may not exist. My comments have always been based on the risk of missing something good rather than the risk of hell. You don’t seem to address that.

    2. If there was zero likelihood, then we wouldn’t be discussing this. The point is that neither of us believe the existence of God can be answered as 100% certain yes or no. So there is some likelihood either of us might be wrong, and it should be considered by both of us.

    ” it is very interesting that a Templeton-funded study gave evidence of harm, and that you’ve chosen to blatantly cherry pick from the results in order to present the best possible picture for your position”

    Thanks for bringing this to my attention. I have already made the required small change to that page. But I am disappointed that after all this discussion, you would accuse me of “cherry picking”, effectively of dishonesty. I try to be scrupulously fair and honest in all my references.

    In this case I think I can see what happened. I tried to check the link I gave on that page, to see what my reference actually said, and the link doesn’t work. So I searched and found this press release from Templeton. And it showed three interesting things.

    1. It had two links, to the study manuscript and the original press release, and neither link now works either. I presume one of them was the link I was referring to and used in commenting.

    2. Nine out of eleven links on that page to press articles say that the study found no effect; none of the quotes say negative effect for those told they were being prayed for. It seems the press got the same message I reported.

    3. The press release seems to be correcting a previous Templeton release on the matter.

    I searched some more and came up with a link to a peer reviewed paper which I am sure I have never seen before.

    So putting all that together, I think I referenced an original Templeton press release which gave the result that I gave – I’m guessing it only compared the two groups who weren’t told if they received prayer, and found that the prayed for group was marginally, but statistically insignificantly, worse. Certainly that is the result I remember.

    So I am certain I did NOT cherry pick, but reported the original Templeton result, and didn’t know it had been updated.

    But none of this is important because that form of prayer (which isn’t how christians actually pray) isn’t relevant and isn’t what I was referring to.

    ”I also disagree with your assertion that religious belief offers any particular benefits in *this* life that cannot be obtained through non-religious pathways. None of the benefits associated with religious observance seem to exceed the benefits available to members of other similar types of non-religious communities – that is, it is the *community and social* aspects of religious observance or other groups that provide benefits to individuals, rather than the religious observance itself.”

    This is what I was talking about, and here I think the science is against you.

    1. The benefits of religion are more than just the community and social aspects. There is too much evidence to go into here, but religious belief and/or practice benefits brain operation quite apart from any social aspects, gives people more reasons to behave ethically, gives people a supreme “cause” to give their life to (the primary factor positive psychology finds important for wellbeing), can give them transcendent experiences that, even if they’re not genuine, give their lives greater meaning, etc. It is true that some, perhaps many of these effects can be achieved in ways other than through religion, but religion provides are much better pathway.

    2. Even if these benefits COULD be achieved without religion, the fact is that they are generally less likely to be. Religious people in fact do slightly better on almost all measures of wellbeing and prosociality. (Ask me and I can give you references by secular scientists on this topic.) This reinforces my statement above that it is easier through religion, as well as sometimes only possible through religion.

    ”You cannot establish that any hazard (and therefore a risk) exists”

    Again you are missing my point. (1) We don’t have to establish that it exists, only that it has a non-zero probability. (2) A risk of missing a good doesn’t require a negative hazard.

    ”modifying your behaviour to account for a risk is one thing, but modifying your beliefs is something else entirely”

    Again, this isn’t what I said. I have said I think it is worth considering other approaches to knowing than what you have adopted. I have not (I hope) suggested anyone should change their beliefs without good reason, but rather I am challenging what you consider to be good reasons, and suggesting a more open approach to the evidence.

    ”the only thing I, as an atheist, cannot do is provide a firm answer to the question “does God exist?”. I can answer *every* other relevant question carefully and thoughtfully”

    Well, I don’t think atheism can answer (satisfactorily) how the world began, why it is so “fine-tuned”, how we can have objective ethics, the reliability of human cognitive systems, consciousness, so may people’s religious experiences (including healings, visions and mystical experiences), the life, teachings and resurrection of Jesus and how religion has such positive effects on people’s minds and lives if it isn’t true.

    Conversely, I don’t think theism/christianity can satisfactorily explain the extent of evil and suffering, and some think it can’t explain the variety of religions and the so-called hiddenness of God (I think it can). That makes the result quite strongly in favour of theism in my thinking.

    ”I can consider new evidence should it be forthcoming, I can change my mind either way if it is justified by that evidence”

    But maybe not while you hold to such an over rationalistic approach that I don’t think you use in all of life, so why use it here?

    I can live my life in an ethically sound and intellectually honest way”

    I don’t question that.

    ”any god who would condemn me for exercising the rational skills he granted me wasn’t worth my time in the first place (being made as I was, I could never have lived up to his requirements to believe in the absence of evidence which would convince a reasonable outsider).”

    But I would say that this discussion suggests the problem isn’t your rationality but your epistemology. And as I say in the next point, you have your biases and assumptions same as we all do, and it may be these that lead you away from God.

    ”I have to ask, do you genuinely think you can do the same? …. do you know what would make *you* change your mind and give up your beliefs? If you don’t, doesn’t that suggest that you are in fact more paralysed by your belief than I am by my lack thereof?”

    I don’t think I can totally say that, any more than I think you can totally. I have lived for about 54 years with this belief and I have a clear set of reasons for it. I imagine you are the same, though presumably with a few less years! I have changed my mind on many aspects of my belief, but not the underlying theism. Again, I imagine you are the same. So we both have biases and assumptions and different ways to approach the question, and these are probably more important than evidence and rationality – after all, we both have fairly similar evidence and probably fairly similar rationality – it looks like it is something else that makes the difference.

    I think it is clear what would lead me to change my mind, and that is if it was demonstrably wrong. But it clearly isn’t demonstrably wrong, and I think the evidence shows it is probably right.

    My comment about paralysed simply referred to waiting for the scientific level of precision that you ask for, when that is clearly inappropriate for many areas of life and arguably inappropriate here. If one adopts such a standard before one is willing to know and act, we’d know and do very little, we’d be agnostic about almost everything and thus “paralysed”.

  42. Hi UnkleE,

    1) Let me clarify again – it doesn’t matter if *you* believe in hell, I think we can both agree that if our concern is about modifying our behaviour and/or beliefs in order to account for risks in the afterlife, then we should focus on the *greatest imaginable hazard* in order to make the *best possible case* for your principle. We can apply your principle to some vague, less good afterlife, but I could simply say that the benefits you’re pointing to aren’t worth the costs in this life – but if we say that hell is on the table, then I should take your principle as seriously as possible. Again, look up “steel man” as the intellectually rigorous alternative to the straw man fallacy – that’s what I’m doing here. Also, if you can’t accept that, then assume that when I refer to hell I’m simply using it as a shorthand for “whatever it is that you believe will happen after we die” – don’t get hung up on the specific word, because my arguments stand up however you want to define it (and are simply *weakest* – although I think still entirely solid – against hell). But, that said, I’ve used “hell (or whatever)” in previous posts to try and make my stance on the matter clear – I won’t continue with that for the sake of (relative) brevity, so please interpret it in the light I’ve just defined.

    2) Again, you’re confusing things that exist with our knowledge of what exists. Hell *does or does not exist* – unless you want to contest the law of excluded middle, of course. We *do not know* whether or not hell exists – which is why we’re arguing about it. The two issues are entirely separate, though – in 2010, the Higgs boson *did exist*, but we *did not know* that it existed – which is why physicists spilled a lot of digital ink arguing about it. Right now, god and hell *do or do not* exist, but we do not know whether or not they do – which is why we’re doing the same.

    3) Cherry picking may or may not be dishonest – it’s in fact a common cognitive problem that’s fairly natural. I’m sorry if it came across as accusing – that was never my intent. But you’d previously pointed me to that page as evidence for one of your claims, and that study was famous (and famously poorly reported) and an easy catch. The study was never updated – that fact that people who knew they were prayed for had worse outcomes was noted in the original conclusion to the article (at PubMed or the full study – Templeton simply cherry picked in their press release, and the media largely took the bait. Media coverage from self-interested press releases is never particularly reliable, unfortunately – not least because few journalists have the skill required to fact check scientific studies. And you’re right – this has little to do with the discussion, and only came up because I was disputing any claims to benefits of religion in this life (beyond anything that is available to non-religious people who want it), and went back to check over that page for anything I’d missed.

    But again, apologies if you thought I was accusing you of dishonesty – on re-reading, I can see how that might have come across, but it wasn’t intended as such (it was simply a surprise to me that you’d have missed that, although I can understand how it would have happened, particularly if you relied on the honesty of Templeton).

    4) As to benefits… I’ve seen many of the studies. They’re generally not high quality, they’re correlative (no longitudinal studies looking at people who end up going from religion to atheism, or from atheism to religion, for example), and they frequently fail to correct for socioeconomic status, heath, social aspects of religious observance, or the cultural and personal biases of people who already see religion as valuable or the only way to obtain the benefits in question (it’s not for nothing that “where do you get your morals?” or “how can your life have meaning?” are such common questions thrown at atheists, after all – even by yourself in this very post!). It’s very easy to get a false positive result (either maliciously or simply through lazy trial design) in *physics*, let alone in the social sciences! We also have a number of studies which demonstrate that benefits amongst religious people depend far more strongly on interpersonal relationships within the congregation than they do upon attendance or piety (if you don’t have strong friendships with other congregants, then your level of religiosity is independent of your wellbeing or happiness).

    Your comment suggests that I wasn’t entirely clear on one point, though. The benefits are not *just* the social and community aspects, but all of the benefits seem to derive *from* the social and community aspects. When you control for other factors, the existence of relationships between congregants is predictive of other benefits, but religiosity is not.

    5) I would strongly argue that all of the benefits (except religiosity itself – which is *only* a benefit if the god you believe in exists and wants you to act as you do) are fully available to atheists, and even anti-theists. Unless you’ve a solution to the Euthyphro dilemma, I certainly *hope* you’re not suggesting that morality is dependent upon god. “Meaning” is a complex issue, because people define it so nebulously (do you mean purpose in life? I got that – from benefiting the ill, taking care of and enjoying my family, engaging in my hobbies, and so on. But do you mean “a sense that everything happens for a reason”, as many seem to? You know – “well, I lost my job, but that gave me time to learn a new skill which I used to land a better one – must have been god’s plan for me all along, even if I couldn’t see it back then!”. Because I don’t have – and do not want – anything like that). Transcendent experiences can be obtained without religion – drugs, meditation, sensory deprivation, and so on are all “easy” ways to get there.

    And *even if* these benefits are *more common* among (because they’re certainly not exclusive to) the religious, why should we ascribe value to religiosity if it’s not the causal factor? If it is simply the result of being part of a strong community, well, why not focus on building strong communities without all the nonsensical baggage attached to religion? Let’s build secular institutions that people want to be involved with – UU and humanist “churches”, for example, or just clubs for people who enjoy a hobby – that seek to accrue the benefits without teaching falsehoods (or, at the least, teaching unknowns as truths).

    Furthermore, there are benefits to non-belief that are at least as valuable as those you’ve pointed to. No afterlife = greater sense of urgency, and thus greater motivation to do the things you care about here and now. No god = taking responsibility for your own life, and not expecting daddy to step in and fix it all up for you. No dogmatic scriptures and rules = better societal outcomes because we can more readily see the world in all its shades of gray, instead of the black-and-white false dichotomies that religion so often puts forth. Lack of dogma also means that we can change in light of new evidence and ideas, instead of being forced to fight against 2,000 year old ethics that are placed on an undeserved pedestal through cherry picking, apologetics, moral relativism, and playing god by attempting to interpret the scriptures in a modern light.

    6) Missing a benefit *is* a hazard. Furthermore, you *cannot* establish that there is a non-zero possibility of hell – as I said at the top, hell *does or does not* exist, thus the possibility of hell is 1 or 0 (there is no in between). We do not *know* whether it exists, so the best we can say is that the best evidence suggests one way or the other. The best evidence for *any* sort of afterlife is, to be blunt, crap; the best evidence for any *particular* afterlife is even worse. Given that there are multiple competing and mutually exclusive hypotheses (Christian hell, what you believe, Islamic hell, just to grab three we’ve discussed already), we cannot justify the claim that *any* of them are possible, let alone decide which one to treat seriously, and we should therefore ignore all of them until someone gets their act together and finds some evidence we can at least get through traffic court.

    7) You say that “it is worth considering other approaches to knowing”. I ask you, what are those approaches and why should I trust them to lead to true conclusions? I’ve also, I think, identified the very lowest bar of evidence that I would accept for a religious claim – John Loftus’ outsider test for faith. Can you point me to a *single* piece of evidence for your belief that would stand up to that test? If not, why should I lower my standard of evidence to a level that you yourself would reject as invalid (because if you’re not going to accept a claim when a Muslim presents it as evidence, why should I accept a substantially similar one from you)?

    8) Firstly, I’d point out that “relevant questions” referred to the ones regarding the existence of and belief about god. But what is a “satisfactory” answer to those other questions? Is “I don’t know” satisfactory? If not, why not – particularly given that it’s the only truthful answer that anyone can give to them (even you, since you have noted that you’re not 100% confident about your beliefs – even assuming confidence correlates with truth)? An atheist worldview can absolutely propose answers to those questions (just a a theistic one can) – but we don’t pretend to certainty when it isn’t warranted. How did the universe begin? Seems to be a big bang followed by inflation – dunno why, we’re working on that and we’ll get back to you if we find more info. Why does it appear fine tuned? Well, it doesn’t – it’s compatible with life, so we’re here to notice that, and because we’re tiny, self-interested little creatures with myriad cognitive biases we *presume* fine tuning without evidence. The reliability of human systems is a consequence of evolutionary adaptation (unreliable systems are selected against). Religious experiences of various types are reducible to well understood psychological phenomena (even if pinpointing the cause of a “specific” experience is difficult or impossible, doesn’t mean we can’t explain the “types” of experiences). I don’t care about Jesus’ story – the ideas he spouted were either specific to his faith (and thus unimportant if it’s not true), or established prior to his time, he offered nothing that was sufficiently revolutionary to be considered impossibly far ahead of his time. And the *fact* that religion provides what benefits it does whilst being false is easily established – albeit not explained – by the fact that multiple mutually exclusive religions offer the same benefits (and studies indicate social factors correlate with and likely explain the broader benefits people experience). I’m not going to argue these points – I’m sure we could go down a lot of rabbit trails – but I fail to see how any of those explanations aren’t at least as satisfactory as “I dunno, but I know someone (god) who does!”.

    9) I’m not going to pretend to be a Vulcan – of course I make irrational decisions, and believe things that are false. But I do my very best to avoid those pitfalls, to think through ideas before accepting them, to examine my biases, to confirm the news I read, to independently corroborate evidence – and particularly on issues which matter greatly. I freely and rapidly change my mind when faced with definitively contradictory evidence – even at the cost of some embarrassment (being rational – I’d rather be embarrassed once and admit that, than be embarrassed again and again and again…). I take shortcuts – I have trusted sources, for example, who I will generally take at face value because they have a solid reputation for fact checking and for correcting their errors, but I question even them (I sent a correction to one of them just last week, in fact!). So no, I’m not a perfect thinker – but I try my best to be a good one, and to apply some basic standards of reason and evidence to everything in my life in order to have my model of reality align as closely with *actual* reality as I can. I try to use an epistemologically valid process for assessing claims and evidence, and I seek to hold all my conclusions as contingent (even if I’m forced to act as if they are unquestionable in order to live my life from time to time).

    That brings up an interesting point, though. We are discussing perhaps the single most important question a human can consider – the existence of god and our fate for eternity – and you want me to *lower* my standard of evidence below that I would apply to my choice of spouse, decisions for my healthcare, or even which car I should buy?!? I’m sorry, but that seems not only completely irrational (our standards of evidence should *increase* as the stakes do), but it actually seems self-serving (unintentionally, I’m sure, but nonetheless – you’re effectively saying that I *should* believe what you believe, even if it means abandoning those epistemological standards which generally work in our favour when applied to any other aspect of our lives). I’m sorry, you can tell me how great the car you’re selling is, but I’m still going to kick the tyres – and the outsider test for faith is the very least amount of tyre-kicking one can do for claims about religion (and is *at least* the standard that you apply to claims made by those outside your own faith – and thus surely the very least to which you should entitle me).

    10) I am not committed to any conclusion about almost any matter (see point 12) – and as such, I can freely change my mind on any of them. You give me evidence and I’ll switch sides – it might take a little while to mull it over, check the sources, look over any criticisms, and so on, but I’ll get there pretty damned quick most of the time. And I’m unlikely to ever believe 100% that god exists (unless we can solve the solipsism problem, at the least) – but I’ll certainly adopt a high level of confidence (at least as high as my confidence in, say, evolution by natural selection, or our ability to interrogate molecular structure with x-rays or magnetic resonance). Of course, I’m sure god himself could get me to 100%, if he actually wanted that… But I am entirely confident in saying that, if presented with valid evidence, I would change my mind to the degree warranted by that evidence. I’ve done it before on matters of deep importance to me, and I’ve no reason to think that my ongoing studies in philosophy and cognitive science, not to mention my attempts to engage with views that oppose my own, are going to make me *less* likely to change my mind when appropriate in the future.

    You, however, have stated that you would need a demonstration that your beliefs are false – but what, exactly, would that look like? How does one demonstrate that a thing – particularly a god who likes to hide – does not exist? I don’t think too many people would argue that you’re wrong to say you’d stop believing if you were *proven* wrong, but if that’s a standard that cannot be met, how is it materially different from simply saying “nothing could convince me”? As an alternative to that, though – you say you think it’s probably true – what would change your mind on *that*? What would it take to get you to 50:50?

    11) “we both have fairly similar evidence and probably fairly similar rationality – it looks like it is something else that makes the difference” I would contend that perhaps it’s our starting assumptions… You start with the assumption that god exists, and that necessarily colours your subsequent arguments (if “circular” can be counted a colour!). But if we start with facts and try to get to god, we can’t do it – because there’s not a single fact that leads us there by logical necessity (I’d say not even as a matter of empirical likelyhood). As I’ve said before, “god” is a hypothesis about how fact X came to be true – but it is all too often treated as a conclusion by people already committed to the idea that god does (or even *can*) exist.

    12) Finally, I would again point out that belief and action are two different things. I would also reiterate that belief and confidence are different things. We can be confident that some thing is true or not, without having to definitively believe it’s actually true. We can (and must) act in many cases on the basis of information we know is incomplete. But none of that is justified in the case of god, because we have no reason to think that god is possible, let alone that he exists or that he wants us to act in specific ways. We also have multiple competing claims, from which we can only choose a very small set, and many of which can lead us very bad places (and at the least cost us *something*).

    And frankly, I *am* agnostic on almost everything. I know some “I” exists, I know my own feelings and thoughts, and I know things that are logical or mathematical truths. But anything outside my own mental world is subject to the problem of hard solipsism, not to mention problems with human senses and cognition, so I am agnostic about them (I do not, when speaking carefully, claim to know any of them). I certainly believe many things to be true – which is to say I have a high degree of confidence (not always justified, as discussed at point 9), but I *know* I could be wrong about almost any of them so I am never 100% confident about anything I believe. This, again, is why it’s important to separate what we believe, what we *actually* know, what we *think* we know, and what is true – all different categories of things that are too often conflated.

    Thanks again, UnkleE.

  43. Hi UnkleE,

    Sorry again about double-posting, but an important point occurred to me.

    When you suggest that I would be averse to changing my mind and believing in god should sufficient evidence present itself, I have to ask, what possible reason would there be for me to *not* change my mind? Like most people, I do not object to the idea of living beyond my own death – various types of afterlife would in fact be great! And if I had a good reason to believe that an afterlife existed, and what I needed to do to get there, why *wouldn’t* I get on board?

    This idea, I think, stems from the commonly held (not necessarily by you) belief that atheists secretly know that god exists, but reject him so that they can sin freely. Obviously that’s an absurd concept, but it also presupposes that atheists are *resistant* to the idea of god, rather than just *unconvinced* by it. Maybe some people do that (although they’re by definition not atheists if they secretly “know” god exists, even if they choose to claim the word), but I hardly think it’s the norm (and we’re both well aware that people can be devoted to god and still commit monstrous acts – meaning there’s no need to pretend to be an atheist in order to sin massively!).

    I am not committed to atheism – I am committed to believing as many true things and as few false things as I can manage, and nothing I’ve ever seen has given me the slightest reason to think “god exists” is true. I would certainly *like* it if some gods and some afterlives were real things, but what I like has no bearing on what is true. I’d rather concern myself with the latter instead of simply accepting comforting delusions, but I will *jump* at the chance to live forever – just as soon as I have a good reason to believe that such a thing is possible, and that I’ve got good information about how to achieve it. Until then, I’m going to live the best life I can and deal with things about which I have no justifiable knowledge and over which I have no control *if* I wake up from my grave.

    All the best, UnkleE.

  44. Hi,

    I find it interesting that you are so interested in using hell as an example, even after I say I am not, because you say it is the “greatest imaginable hazard” – and the dismiss it! So let’s stick with the argument I make, which is that believing in God may confer benefits in the next life and certainly does confer some small benefits in this life. And all I am saying is that the nature of the question (is there a God?) suggests to me that a strict “scientific” approach is no more suitable than it is in other areas of life that have uncertain evidence, unquantifiable benefits and costs and the possibility that waiting to make a decision may be counter productive. I don’t think you’ve fully responded to that, but if you don’t think there is any truth in what I say, we can count that part of the discussion as finished and unresolved.

    Thanks for your comments on Templeton. Yes, I think I did rely on their press release. The correction doesn’t actually change anything material. So I’ll leave it there.

    You say: ”When you control for other factors, the existence of relationships between congregants is predictive of other benefits, but religiosity is not.”

    You make this statement with confidence – on what is it based? My statements are based on these summary studies (and much more as well).

    Would the World Be Better Off Without Religion? A Skeptic’s Guide to the Debate: ”the data consistently point to a negative association between religiosity and criminal behavior and a positive association between religiosity and prosocial behavior. Both relations are modest in magnitude and ambiguous with respect to causation. At the same time, they cannot be ignored by partisans on either side of the discussion.”

    Religious Prosociality: Personal, Cognitive, and Social Factors This study recognises the connection between religion and other social factors, but also discussing priming studies where these factors are not included, and says: ”We also reviewed evidence from a wide variety of studies that religious cognition can increase prosocial behavior, (e.g., cooperation, sharing, and giving). These studies often use priming techniques to activate religious cognition, and though it is not entirely certain what kind of ―religious cognition‖ is activated (i.e. the religious constructs being activated by words such as ‘God’ or ‘church’), the preponderance of positive findings suggest that these concepts are tightly connected to prosocial concerns.”

    Different Effects of Religion and God on Prosociality With the Ingroup and Outgroup. Another review of priming studies that don’t have effects from social factors, and says: ”we found that religion primes enhanced prosociality toward ingroup members, consistent with ingroup affiliation, whereas, God primes enhanced prosociality toward outgroup members”

    Comments aren’t for atheist evangelism. Connor Wood has done a PhD in the scientific study of religion, and he says: ”the data that religion has social and individual benefits is so overwhelming that saying that religion has no benefits is active science denial.”

    Why Your Brain Needs God Andrew Newberg and Mark Waldman say: ”faith is the most important thing a person needs to maintain a neurologically healthy brain. …. By faith, we mean the ability to consciously and repetitively hold an optimistic vision of a positive future — about yourself, and about the world. When you do this — through meditation, prayer, or intensely focusing on a positive goal — you strengthen a unique circuit in your brain that improves memory and cognition, reduces anxiety and depression, and enhances social awareness and empathy toward others. And it doesn’t matter whether the meditations are religious or secular.”

    So often the good effects can be achieved by non-religious means, but then they add: ”However, when meditation is religious and strengthens your spiritual beliefs, then there is a synergistic effect that can be even better.”

    You say: ”Furthermore, there are benefits to non-belief that are at least as valuable as those you’ve pointed to……”

    I’m sorry, but these are just assertions, backed by no data I know of and contradicted by much data including what I’ve quoted above. And they show little understanding of christianity. Positive forms of christianity are shown by evidence to increase motivation in many cases.

    I think this comment has become long enough, so I will stop there and comment on the rest of your comments later. Thanks.

  45. Hi UnkleE,

    My “focus” on hell was a response to your previously espoused principle that we should modify our behaviour in the light of risks – for example, “the risk factor makes it smarter to lean on the side of believing or obeying” and “belief in God can have some clear advantages in this life and the next“. If we’re talking about modifying behaviour in light of risks in the next life, then the obvious place to start is with the greatest risk we can imagine – because that’s where your principle will be strongest (and my counterarguments most tested – because if I can reject your principle when hell is on the table, I can reject it for *any* lesser risk). Furthermore, I stated that I was happy for you to accept “hell” as my shorthand for “whatever afterlife you believe in”. Your original principle was at least in part appealing to consequences (positive or negative) in the next life, and my comments on hell were directed against that portion of your claims. At this point, I have to ask – are you walking away from that principle (as applied to unknowable claims like the existence of or rules for attaining an afterlife), or do you still think it’s valid but are unwilling to defend it? I have no problems with setting it aside, but I’d like you to at least clarify for me where you stand on that because we did devote some time to the topic.

    You comment about the applicability of a scientific approach to the question here. I’d like to clarify – is the question you’re referring to “does god exist”, “should we believe that god exists” or “are there benefits to belief in god”? Your subsequent comments suggest that you *are* referring to the second question – but that’s *not* a scientific question! The first and third are scientific questions – we can, in principle, answer them through scientific inquiry, and I see no reason to assume that we should lower our standards for interrogating the universe on such an important question (and therefore see no reason to come to *any* conclusion on the basis of extremely poor evidence – evidence which, in every case, fails even to pass the very minimal standard of the outsider test). The second, though, is entirely philosophical, and we cannot address it in a scientific manner at all. We should absolutely rely on scientific evidence when evaluating the question, but we cannot answer it in a scientific way (“should” is usually a good indicator that we’re dealing with an non-scientific question).

    But I think I *have* responded rather fully to *that* question. When we’re making rational decisions, unknown unknowns are poor reasons on which to base choices. We can consider them occasionally, but they should be given very low priority most of the time. To go back to one of your own examples – when we consider licensing drugs, we look at knowns (what are the benefits and risks detected in studies), and we look at known unknowns (we specifically test drugs for certain side effects that we know can arise even if there’s been no evidence of the drug causing those things), and we license or not on the basis of that information. We do *not* appeal to unknown unknowns (although we try to account for them through post-market surveillance for unexpected benefits or side effects), because doing so would, as you put it, paralyse us – we can posit all sorts of things that might happen, or might have slipped through the tests, but we cannot justify decisions on the basis of what *might* be true, we have to act in accordance with what the best evidence tells us is most likely to be true. Of course that leads to bad things happening – but it would be far, far worse if we let fears or hopes override best available evidence (we would license *everything* on the off chance it could help, or *nothing* on the basis that it might kill someone). The *only* time I can see unknown unknowns as worth any particular weight is when you’re dealing with a seriously borderline case (mixed evidence both ways, for example), and where you can make a strong argument that it is more likely that there are, say, negative unknowns than positive ones (to drugs, again – there are more ways to hurt a person with a drug than there are to help them, so if there’s limited evidence of benefit *or* harm, we should err to assuming that the drug will be harmful).

    The afterlife is not at all a case like this, however. With drugs, we *know* unexpected negative side effects can occur; with the afterlife, we have no rational basis for claiming that it is *possible*, let alone that it does in fact exist. We have no positive evidence which is reliably suggestive of an afterlife. There is nothing that we *do* know of that is remotely analogous to an afterlife from which we can reasonably extrapolate. So appealing to an unknown unknown (and, specifically, a *particular* unknown unknown amongst *many similar ones*) is entirely irrational and not compelling to someone who doesn’t already believe what you do (if I don’t believe that there’s an afterlife, then you can’t say that I should believe in an afterlife in order to maximise my chances of accessing an afterlife – if I think the likelihood of a benefit is zero, then the magnitude of the benefit doesn’t matter!).

    As to your criticism of my assertion, I would state again that I never argued that religion had *no* benefits – I stated that it had *no benefits which could not be achieved through non-religious means*. The Lilienfeld piece you linked cites three commonly proposed (but admittedly poorly supported) hypotheses to explain the benefits of religion (I’ll paraphrase) – “(a) the so-called “hellfire hypothesis”; (b) religious beliefs bind individuals more closely to communities, families, and others (social control theory); and (c) religious beliefs foster shame and guilt regarding unethical actions, thereby deterring people from engaging in them (rational choice theory)”. None of these things are unable to be achieved to a similar degree via non-secular means (e.g., we use surveillance and punishment to deter crime, and the rest are purely social mechanisms anyway). That there may be a difference between secular and religious people *now* does not mean that religion is the best way to foster the desireable traits – it may simply be that religion has a few millenia head-start on secularism (and demonised it thoroughly during that time to boot), and we’re just now catching up in terms of offering the secular social alternatives that people need and benefit from. Charity might in fact be the best example of this – giving in church is not just easy, it’s *expected* in many religious communities, so doing charity is simply a part of Sunday morning. There’s no secular equivalent to this, so it’s not surprising that religious people would read as more charitable, when in reality it may simply be the case that there are less barriers to and more *social pressure* in favour of being charitable! This is kind of supported by evidence which shows that, when you discount donations to churches and similar groups (i.e. donations that are specifically or predominantly for the promotion and maintenance of religion), secular people tend to be at least as charitable as their religious counterparts.

    I also noted that the majority of studies – perhaps all of them – are correlative and have little or no power to identify causation. For example, does religious belief make one more self-disciplined, are self-disciplined people attracted to religion, or do people with inherently poor self-discipline simply drop out of religious organisations which demand it? The *only* way to distinguish those possibilities is through a longitudinal study which measures self-discipline regularly over a period of time, looking in particular at changes that take place when one goes from strong belief to atheism, for example – and I’ve not seen that type of study done!

    If we’re going to appeal to correlations – the most secular nations in the world also tend to find themselves toward the top of wellbeing indexes. Secular charities tend to rank higher than religious charities in terms of percentage of donations that go to external purposes (i.e. to not paying for the operation of the organisation itself). The most religious states in the US tend to be the ones with the highest crime rates – and the states with the lowest crime rates tend to be the least religious. And atheists are *seriously* (by an order of magnitude) underrepresented amongst prisoners (again, in the US). I’m sure I could find other examples which demonstrate solid positives associated with (but not necessarily caused by) lack of belief – all of which are neither better nor worse than your own claims, but do support the argument that religion is not superior to secularism.

    I would also point out that one of your own citations points to studies which show that the social factors associated with religion are predictive of religiosity (the first Preston paper, on p 22, itself citing Koenig et al., 1997), which is rather an interesting piece of support for my assertion that I’d not come across before (if social factors determine *religiosity itself*, then they ultimately determine all the other benefits associated with religion). I’ll have to try and get my hands on the original research.

    But to your specific citations:

    1) This is a long one, and I’m not going to offer much comment. Suffice it to say I *like* the way Lilienfeld goes about presenting his ideas (not just here – he’s someone I’ve encountered before), and I don’t disagree with much of what the authors have to say here. Many atheists *do* go too far in claiming that religion offers no benefits, which is demonstrably false. I stopped at “no benefits that aren’t available through secular means” – which *does* mean that I hold religion as a net negative influence on the world (if it offers no extra benefits, but introduces new harms, then it’s a net bad), whilst freely acknowledging that it can have benefits, particularly at the individual level. They also highlight my issue with purely correlative studies, and that “religion” is a terrible category (given the thousands that exist in the world, all with vastly different tennents and thus likely different effects on behaviour). All in all, I don’t think that they’ve refuted any of my claims in this article.

    2) Correlation versus causation – does religion encourage prosociality, or are prosocial people attracted to religion? One of my philosophy of religion lecturers (who is, I’ve recently discovered, rather world renowned) told us several stories about one of *his* teachers, an atheist priest who stayed with the church after losing his faith purely because he felt he was able to do good there. So, he was contributing to the prosocial appearance of the religion, whilst doing nothing he couldn’t have done if a similar secular structure was in place. Not saying this is anything more than an anecdote – but it’s a factor the article doesn’t account for, and one which is common in many ways (check out the “clergy project” – a support network for religious leaders who have discarded their beliefs but, for whatever reason, can’t or don’t want to leave their organisations).

    Furthermore, the article itself points out the complex relationships between religiosity and pro-social behaviour. The *type* of belief matters (e.g. intrinsic vs. extrinsic vs. quest religiosity), affecting the actions one takes, and who one attempts to help – with quest religiosity being the most beneficial, and the most similar to a secular person who simply wants to help people for the sake of helping (as opposed to doing it for personal (extrinsic) or social (intrinsic) reasons). Fundamentalism – the *strongest* form of religiosity – results in less helping of outgroups (i.e. everyone not like you), which seems to be an example of greater religiosity leading to more harm than good – and supportive of my point that religiosity isn’t easily linked to benefit. And overall, they note that no matter your motivating religiosity, all religious people were more interested in helping ingroups and less so in helping members of outgroups – hardly a socially beneficial outcome.

    They note that there is no obvious difference in how much religious or non-religious people help others. They state that there is no clear evidence that the salience of one’s beliefs has any effect on prosocial behaviour. And they even note that the monopoly that religions have (I would argue dishonestly) claimed over moral thinking may lead to religion being associated with morality in the minds of people – therefore explaining the priming effects seen in many studies (suggesting that if we simply started teaching secular ethics, instead of pretending that religion is the only source of morality, even the priming effect would be reduced).

    Again, we’re left with equivocal evidence (perhaps slightly positive) that is not at all inconsistent with social effects or benefits that can be obtained without religion.

    3) This study is of no value to the discussion at hand – it seeks to examine the effects of priming with thoughts about religion versus thoughts about god. It does not, for the most part, examine a non-religious control, and most of the results found that the effects of priming were independent of the subject’s personal beliefs about god or their religiosity. Also, I’d note that priming about religion resulted in greater favouring of the person’s ingroup – hardly a social benefit. And as an atheist, if asked the question I might well answer precisely as they found – a god would likely prefer I help people who most need it (which are going to be people *not* like me more often than not, give my relatively privleged circumstances), whilst religious leaders have an incentive (even a responsibility) to provide support for their own congregants over others, all things being equal (and that’s ignoring the cynical little voice I hear screaming about Catholic property ownership as evidence of pure self-interest in that church!).

    4) Wood says that there are benefits – and I never said there were none, just that they do not seem to stem from *religion* but from *a community formed because of shared beliefs*. If we build strong communities, we don’t need magical thinking to get the benefits. This is exactly what I was claiming.

    5) These guys are framing reasonably well understood psychological mechanisms like rationalisation, cognitive dissonance, optimism and so on as “faith”, which is hardly a reasonable equivocation (we have a word for what they describe – irrational. We don’t need to use the baggage-laden “faith” here). Furthermore, the quote you’ve provided indicates that meditation and focusing on a goal (neither of which are necessarily associated with religion) are mechanisms by which the benefits they speak of can be obtained. Again – I never said “no benefits”, just “no benefits that aren’t available in the absence of religion”. As to the proposed synergism of actions which build upon religious belief – yuh, big surprise, when you basically tell yourself you’re right (and do things to make you feel even more right) you feel better about yourself. That’s called confirmation bias. But happy delusions (and poor critical thinking) are not necessarily a personal (let alone social) good overall, though!

    As to the benefits of non-belief – first off, you don’t get to simply dismiss self-reported benefits, unless you’re willing to jettison much of the research into benefits of religious belief (and your some of own claims). Some is empirically verifiable – like the prosocial behaviour research we’ve just looked at – but much is purely self-reported – including “giving people more reasons to behave ethically, giv[ing] people a supreme “cause” to give their life to … giv[ing] them transcendent experiences that, even if they’re not genuine, give their lives greater meaning” which you pointed to! Second, I never said one or other was superior – certainly not as a blanket statement, given the vastly different things that different people believe, and the reasons for which they do so. But you pointed to entirely self-reported benefits of belief, so I pointed to self-reported benefits of non-belief for the purpose of showing that it’s not reasonable to argue that one is obviously superior to the other. I’m not pretending that these are definitive, or of large magnitude, or even superior to the benefits that religious people (yourself included) claim – just that they are at least as valid and therefore worthy of equal respect when considering the overall value of belief or non-belief.

    Finally, let me throw out a handful of articles that support my point.

    1) Galen, 2015: “organized nonbelief is also associated with well-being and prosociality equivalent to that seen with organized religious belief. Notable areas of relative advantage for nonbelievers are in the domains of outgroup tolerance and moral universalism.”

    2) Lim and Putnam, 2010:“this study offers strong evidence for social and participatory mechanisms shaping religion’s impact on life satisfaction. Our findings suggest that religious people are more satisfied with their lives because they regularly attend religious services and build social networks in their congregations… We find little evidence that other private or subjective aspects of religiosity affect life satisfaction independent of attendance and congregational friendship.” I note that religiosity is important *within* the religious group (i.e. if you’re not sufficiently religious, you don’t get the benefits of belonging to that group – which is not unexpected, given that religious groups are primarily about shared belief, and if you don’t share the belief it’s not surprising if you don’t fit in), but it is not predictive of happiness otherwise.

    3) Decety et al., 2015: “religiousness was inversely predictive of children’s altruism and positively correlated with their punitive tendencies. Together these results reveal the similarity across countries in how religion negatively influences children’s altruism, challenging the view that religiosity facilitates prosocial behavior.”

    4) Ellis et al., 2012: “[T]here is a substantial positive correlation between most religiosity measures and fear of death… on average females were more religious and feared death more than did males, and Muslims expressed considerably greater fear than did members of any other major religion.”

    5) Zuckerman, 2009: “Atheism and secularity have many positive correlates, such as higher levels of education and verbal ability, lower levels of prejudice, ethnocentrism, racism, and homophobia, greater support for women’s equality, child-rearing that promotes independent thinking and an absence of corporal punishment, etc. And at the societal level, with the important exception of suicide, states and nations with a higher proportion of secular people fare markedly better than those with a higher proportion of religious people.”

    And I’m sure I could go on, but my point stands unrefuted – I’ve yet to see evidence of a benefit that is available to religious people that is not available through secular means, and there is insufficient evidence to support the claim that religion is necessarily more efficient in providing those benefits than secular organisations (even if the secular world is having to play catch-up and beat back religious propaganda in order to establish those benefits).

    Let’s be clear here. You have not presented any evidence that *religion* is the source of the benefits we observe, as opposed to *something that is a part of religion* that we could isolate and exploit in order to get the benefits without the magic. I have presented research which isolates social factors as being more important than strength of belief; which shows that non-believers are equal to believers when both have similarly strong social networks amongst the groups they associate with; and which shows that the commonly claimed benefits of belief are not so clear cut as believers might like to pretend (and the truth may in fact be the opposite of common wisdom). Again, I would lay the burden of proof squarely on your shoulders to demonstrate that religion provides benefits *above and beyond* an organised social group, because that’s the positive claim here – and just because it’s the commonly held view, doesn’t mean it’s not the one that needs to be justified. The references you cited do not achieve that goal.

    Finally, to tie up the last few posts, let me try to distil out what I think are my most important questions I have for you:

    1) Why do you think I would be resistant to adopting a belief in god if I were presented with epistemologically sound evidence for his existence?
    1a) What evidence would get you to the point of complete uncertainty (i.e. 50:50 on the existence of god)?
    1b) What would convince you to discard your beliefs as unfounded (without necessarily rejecting the possibility that god might exist – which is the position that I am in, as an agnostic atheist)?

    2) Why do you think I should be willing to adopt a standard of evidence which is lower than the one you yourself apply to religious claims?
    2b) Is there a single piece of evidence that you can offer which passes the outsider test for faith?

    3) Do you stand by your uncertainty-based principle of believing (or acting as if you believe) in order to reap certain benefits in the next life?
    3a) If so, why are *you* not convinced to believe in (and act to avoid) hell, given that many hell-fearing Christians (or Muslims) would suggest that *you* should apply the very same principle to *their* belief?

    4) Assuming religions are based on a false belief (which most of them *must* be, even if one is right), can you name a benefit of religion that is not available to a non-religious person?
    4a) Can you point to a benefit which is *less* available to non-religious people, where the difference *cannot* be explained by confounding factors – in particular, a lack of equivalent secular structures?

    5) What other “approaches to knowing” would you suggest are valid and reliable pathways to true beliefs, beyond the evidence- and logic-based processes of scientific and rational inquiry? Why is it anything more than special pleading to argue that we should adopt them for this particular question, if they are not at least as reliable as those methods?

    6) Is “I don’t know” a “satisfactory” answer (given that it’s true in most cases) to the questions of cosmology, biogenesis, etc.? If not, why not?
    6a) Is “god did it” substantially different to “I don’t know, but I know someone who does”?

    Thanks, UnkleE.

  46. Hi Cheomit, I’m feeling like we need to review where we’re going here. Pleasant as the conversation has been, I think comments are getting very long and the discussion doesn’t seem to be heading to any conclusion. So I am going to identify what I think are the issues we have before us and work through them one at a time by making one statement on each, seeing what you have to say, and deciding whether we agree on anything, are unlikely to agree on things, or if there is some aspect still worth discussing. And then moving on to the next one. I hope this is OK with you.

    I think the following topics need to be addressed, and cover all your unanswered questions:

    1. Does religious belief truly lead to some personal or societal benefits that non-belief does not?
    2. Should we adopt a “scientific” epistemology to determine whether we can believe in the existence of God?
    3. Do either of us make assumptions or start with a preference for belief or unbelief?

    So I will begin in this comment with #1.

    Firstly, your comments on the studies I quoted. It is interesting that although the studies make some quite definite statements, you say you agree with them while at the same time dismissing the conclusions that don’t fit your own view. To pick three examples:

    (i) Lilienfeld and Ammirati say: “the data consistently point to …. a positive association between religiosity and prosocial behavior” whereas you say: “I hold religion as a net negative influence on the world (if it offers no extra benefits”.

    (ii) You summarise Wood as saying “just that they do not seem to stem from *religion* but from *a community formed because of shared beliefs*”, based on that one reference (?), whereas he has done a PhD in the topic, and over the years I have followed his blog has reviewed scores of references that lie behind his statement: “saying that religion has no benefits is active science denial”.

    (iii) Newberg and Waldman say (my emphasis): “However, when meditation is religious and strengthens your spiritual beliefs, then there is a synergistic effect that can be even better.” But you say: “yuh, big surprise, when you basically tell yourself you’re right (and do things to make you feel even more right) you feel better about yourself. That’s called confirmation bias. But happy delusions (and poor critical thinking) are not necessarily a personal (let alone social) good overall, though!” So you are contradicting these two eminent neuroscientists that religious meditation etc is “even better” than non-religious.

    So you are not accepting the evidence, but distorting it to suit your preconceived view.

    Now we come to your references, and in contrast, I will accept what they say without quibble, as the conclusions of that review, to be included in the mix with other reviews. For example, Galen’s study is not only a review of the science, but a disagreement with the scientific conclusions of McBayer. So we have one study on each side of our question.

    Further, I note the following statements by Galen (my emphasis):

    “there is an extensive history of studies concerning the putative benefits of religiosity”
    “Most of the literature …. has indeed found that religious attendance is related to lower depression, lower divorce risk, greater charitable giving and community volunteering. Likewise the “actively religious” have better physical health and longevity …. these findings refer to religious activity and attendance rather than privately-held belief in God”
    many of the same factors that promote well-being in the religious are also effectively utilized by the nonreligious”

    So even while making the points he does, Galen affirms that (1) his view is not the majority view, and (2) organised non-belief can compete with religiosity on some (not all) areas of prosociality.

    Lim and Putnam make the same admission: “the positive association between religiosity and life satisfaction is well documented”.

    I couldn’t link to the last study and I have no comment on the remaining two.

    Next, a few other studies to throw into the mix:

    Mental Disorders, Religion and Spirituality 1990 to 2010: A Systematic Evidence-Based Review: ”There is good evidence that religious involvement is correlated with better mental health in the areas of depression, substance abuse, and suicide; some evidence in stress-related disorders and dementia; insufficient evidence in bipolar disorder and schizophrenia, and no data in many other mental disorders.”

    Greenfield & Marks: Religious Social Identity as an Explanatory Factor for Associations between More Frequent Formal Religious Participation and Psychological Well-Being: ”more frequent formal religious participation is associated with having a stronger religious social identity and that this aspect of identity, in turn, accounts for associations between more frequent formal religious participation and higher levels of subjective psychological well-being”. This seems to be saying that religious participation leads to social identity which then leads to better psychological wellbeing.

    Does Religion Make You Healthier and Longer Lived? Evidence for Germany: ”It was also clear that religious belief and church attendance are associated with health-protective behaviors and attitudes, including taking more exercise, not smoking and higher life satisfaction” So the good effects of religion aren’t just confined to USA.

    Want to study the science of religion? Start with this MOOC. Connor Wood outlines some basic ideas in the scientific study of religion, including the complex nature of conclusions, but also: “religious ideas or symbols “prime” people to behave in certain ways” and “religion influenced the development of cultures over time”.

    So here now is a summary of all this evidence (as I see it):

    1. The relationship between religiosity and wellbeing & prosociality is complex and multifaceted. There are many different results, and no single study can address all the factors. Our conclusions must be based on the sum of all the studies.

    2. Virtually all studies, including the ones you referenced, agree that most studies find a positive correlation in general, and for the majority of aspects investigated.

    3. Nevertheless, there are some aspects where religiosity is negatively correlated with aspects of prosociality.

    4. Some of the positive aspects of religiosity can be attained by non-religious persons if they follow certain practices (Newberg & Waldman) or form organised social groupings. But Newberg & Waldman says religious belief enhances those practices, and Galen agrees that not all of the benefits can be shown to be available to the non-religious. Further, studies universally show that whatever might be available to non-believers, they mostly don’t avail themselves of those opportunities sufficient to change the universal conclusion that in practice and in more cases than not, religious belief is associated with better wellbeing and prosociality.

    5. It isn’t always clear which way the causation goes, but many studies assume or conclude that religiosity is a cause of wellbeing and prosociality. The priming studies and Greenfield and Marks appear to demonstrate this is generally the causal direction.

    6. It is clear that the social relationships are often an important factor, and that in many cases other groups having strong social relationships can gain similar benefits. But Greenfield and Marks show that religious practices generally lead to the social relationships, while the priming studies, the Connor Wood summary above shows that often (mostly?) religion is the cause, while Newberg & Waldman and other studies show that factors other than social relationships are also important.

    7. Questions (raised by Galen) about whether atheism “performs” better than other forms of non-belief, and whether it is belief in God or religious practices that are associated with benefits are irrelevant to my claims. The claims have always been general – religiosity, or religious beliefs and practices, lead to greater wellbeing and prosociality than does unbelief. We can subdivide both belief and unbelief and find better and worse aspects of both, and that is interesting and helpful, but doesn’t change the general statement.

    So that is my summary of the totality of the evidence. What do you agree with and what do you disagree with?

  47. Hi UnkleE,

    I agree that our comments are expanding – but that’s the problem when multiple claims are made that require detailed responses. It’s very easy to throw a pile of claims against the wall – it’s much harder to adequately justify them, or to respond clearly (with evidence as required) to them.

    Your response to my views on the studies again misses my point. I absolutely agree that religions offer benefits to their followers, whilst still maintaining that the overall effects are negative. No cherry picking or cognitive dissonance required – it’s simple logic. The benefits of religion do not appear to exceed the benefits that one can obtain from non-religious organised groups, religion introduces harms to the world, therefore religion is a net harm (if otherwise equivalent religious and secular groups offer equal benefits, but religion also contributes to Islamic terrorism or bombing abortion clinics, then religion is inferior to secularism). Religion could even, say, offer superior outcomes in terms of prosocial behaviours (which I do not think the studies support – good, yes; superior, no) whilst still being a net negative if it has a sufficiently harmful influence in another area.

    Also, again, I’m not arguing there are *no benefits*, I’m saying there is insufficient evidence to support the claim that religion offers benefits *above and beyond* those available to non-religious people – excepting to the extent that religion as a social organisation has a head start on equivalent secular organisations. I’m absolutely not denying the balance of scientific evidence, which is certainly consistent with the hypothesis that religion offers benefits of various sorts – I am simply rejecting as unproven the claim that it offers *greater* benefits when you control for variables that many studies in the areas seem to ignore.

    And as to Newberg and Waldman – I concede freely that I’ve not read their book, but the piece you linked to doesn’t seem to say what you represent it as. They’ve had to redefine faith from the beginning, which clouds the whole issue. Furthermore, they talk about spiritual more than religious beliefs being synergistic with meditation – and since people have spiritual experiences in multiple religions and none, that’s again not surprising under my model. All that they seem to be saying in that piece is that if you focus on the things you find most important and inspiring when you meditate, it works better than when you don’t. If you have primary literature on the topic, I’d be happy to read it, but nothing in the linked piece conflicts with my claims.

    I am quite comfortable in stating that I do not see any place where I’m distorting evidence whatsoever. I can perhaps understand where you see me having a problem with Newberg (and if I saw the original research, I accept that there may be more to their comments than is expressed in the piece you linked) but I cannot see a conflict at present.

    In responding to my references, you note that “we have one study on each side of our question“. But simply counting references isn’t how science works. I’m defending the null hypothesis – religion is no more effective than an equivalent secular social organisation; you, in contrast, are apparently claiming that religion offers benefits *above and beyond* what any secular situation can offer. The balance of evidence needs to be in your favour before it becomes reasonable to accept your claim.

    Also, the minority view is not equivalent to the view of the minority of published studies. If you’re going to rely on the primary literature to make your case, you need to understand the primary literature and its limitations. It’s not for nothing that “Why Most Published Research Findings Are False” by John Ioannidis is so widely cited. The literature is heavily biased towards surprising positive results – which means it’s highly biased toward false positives. It’s also littered with flawed studies, driven by low-quality methodology, unconscious researcher bias, poor statistical techniques, and more than a little outright fraud. It happens to be the best thing we’ve got – but it takes a lot of dedicated skill to teasing out the signal from the noise in the *hard* sciences (where there is a single right answer to many questions), let alone in social sciences!

    Consider homeopathy as an example. I hope we can agree that diluting a substance which causes the symptoms you want to fix until there you’re unlikely to find a molecule of the original material *in a solution the volume of the entire known universe* doesn’t seem like a reasonable way to treat diseases. Despite this, there are *many* studies which find homeopathic remedies to be effective for treating various diseases – in fact, there are probably more studies showing a positive effect than there are negative ones. But because the high-quality studies are negative, we are far more justified in accepting *their* conclusions than we are in taking on the “majority” view, even ignoring the absurdky low prior probability of homeopathy working!

    Also, “many” does not mean conceding some advantage to religion – it’s simply saying that we haven’t looked closely at *everything* yet, or that we might be missing something, or that he’s not 100% sure he’s caught the entire literature. Good scientists (and certainly philosophers) equivocate – particularly in our technical literature. We try to limit our claims to the evidence available, and we point out that we don’t know everything and we could in fact be wrong.

    And, again on your quote from Lim and Putnam – “the positive association between religiosity and life satisfaction is well documented”. *YES, I KNOW*. There are benefits to being a member of a religious organisation (and to being religious if you are). But that does not mean that the benefits go *beyond* those available to the secular – to quote from Galen, for example, “many [studies] have indeed found that strong belief either way is associated with mental health, with affirmatively nonreligious individuals and atheists exhibiting mental health equivalent to the highly religious” (followed by nine citations), or “although … atheists report lower levels of meaning in life … atheists’ lower general sense of meaning does not affect their overall happiness or life satisfaction”.

    I’m not sure why the Zuckerman article failed to link – it’s here if you’re interested (

    I’ll try to be brief in commenting on your latest references:

    1) Yes. And studies show that *involvement in secular organisations* is just as beneficial once you control for confounding factors. Some studies linked in Galen’s piece show a *negative* effect of religion on mental health outcomes in at least some situations (particularly for the weakly devout, as opposed to atheists or the strongly devout who seem about equivalent). Completely consistent with my views – if there’s no additional effect, we expect a mix of positive and negative studies (skewed toward the positive in the literature by publication bias and the file drawer effect).

    2) No, they’re *starting* from the premise that social identity determines benefits gained from religion, not the other way around. “The results provide evidence that more frequent formal religious participation is associated with having a stronger religious social identity and that this aspect of identity … accounts for associations between more frequent formal religious participation and higher levels of subjective psychological well-being (emphasis mine). In other words, your social identity accounts for *both* increased attendance *and* subjective wellbeing – that is, religion doesn’t determine wellbeing, *both* of them are actually determined by a third factor (social identity). It’s the very epitome of the skeptical mantra “correlation does not equal causation”.

    3) Why would it be geographically limited? Unless one religion happens to be true (and geographically isolated), of course. People are people, and they generally benefit from the same sorts of things. Again, though, this simply points out that there are benefits to religion – not that there are benefits above and beyond those deriving from the social aspects.

    4) Again, just from the post you’ve linked, he’s talking only about the fact that religion offers some benefits. Which, again, I am not disputing. Becoming a paraplegic can offer benefits (e.g. resilience), but that doesn’t mean it’s a good thing overall, or that there aren’t other ways to get those benefits.

    So, then, to your summary.

    0) It seems fairly clear that religion offers benefits to people. That’s not a problem for me but, to be blunt, you seem to keep skipping right over the distinction I’m making between “no benefits” (not my claim) and “no additional benefits”.

    1) Yes, it is complex. But given that the null hypothesis is “religion offers no specific benefits above and beyond any other organised social interaction”, “complex and multifaceted” doesn’t cut it – in order to be justified in believing that religion *does* offer greater benefits, you need clear evidence of additional benefits, and that doesn’t seem to exist. We don’t say a drug works because people get better after taking it – we *only* say it works if more people who take it get better than do equivalent people who take a placebo. Given the evidence available, there’s no compelling reason to think that all the extras tacked on to religion are anything more than placebo window dressings on the beneficial social organisation.

    2) It doesn’t matter what “virtually all” studies say. What matters is the balance of evidence, weighted heavily toward the highest quality studies. Nonetheless, the evidence that there are benefits is solid – the evidence of *additional* benefits is not.

    3) Agreed. All we need do is look at terrorists, parents who kick their gay children out of home, or faith healings to prove that beyond any reasonable doubt.

    4) That’s not a reasonable interpretation of Galen – he simply concedes that we haven’t looked at everything yet (rather like far too many doctors shrug off homeopathy with “well, science doesn’t know everything”). Newberg and Waldman, I’ll remain neutral on for now – although nothing they offered in your link was inconsistent with obtaining those synergistic benefits secularly (e.g. through reverence of nature instead of reverence of god) – Galen talks a little about this (and cites to primary literature), in fact. And whether or not non-believers look to take advantage of various benefits has no bearing on their existence, and such a simplistic statement almost certainly fails to account for a huge number of confounding factors (e.g. demographics, social pressure [hard to avail yourself of an atheistic organisation if doing so will result in your religious parents kicking you out], entrenchment of religion over millennia, etc.).

    5) Sorry, but those conclusions aren’t reliable. Supporting causation is *hard*, and I’ve not seen a study with the power to do so reliably. The priming studies have also been addressed – and they’re not particularly surprising to me either. You tell people big daddy in the sky is watching you, they behave; but they also do so if you tell them there’s surveillance cameras in the area, or if you paint a big pair of eyes on the wall. I’m not going to comment further on the priming studies, because they’re underwhelming, they often don’t say what people think they say (as I’ve pointed out previously for the Preston and Ritter study), and they are entirely consistent with my own understanding of the topic.

    6) I disagree, as I’ve outlined. Greenfield shows the opposite to what you’ve claimed; Wood showed no causation; and Newberg (and related studies) seemed to be more focused on *reverence* than *religion* – and there’s no reason that reverence can’t be secular.

    7) Again, I disagree with your claim, and I restate that the burden of proof to show that there are benefits to belief *above and beyond* those available to non-believers (controlling for appropriate factors including socioeconomic status and the entrenched nature of religion), because your claim is an attempt to reject my null hypothesis.

    As an attempt to keep things focused, I’d like to point back to the “important questions” from my previous post – none of which I think you’ve really answered. I’m willing to drop question 4, because that’s what we’ve been discussing (and I think the answer is “no”, because the evidence you’ve offered was not compelling, or even inconsistent with my claims of no additional effects).

    1) Why do you think I would be resistant to adopting a belief in god if I were presented with epistemologically sound evidence for his existence?
    1a) What evidence would get you to the point of complete uncertainty (i.e. 50:50 on the existence of god)?
    1b) What would convince you to discard your beliefs as unfounded (without necessarily rejecting the possibility that god might exist – which is the position that I am in, as an agnostic atheist)?

    2) Why do you think I should be willing to adopt a standard of evidence which is lower than the one you yourself apply to religious claims?
    2b) Is there a single piece of evidence that you can offer which passes the outsider test for faith?

    3) Do you stand by your uncertainty-based principle of believing (or acting as if you believe) in order to reap certain benefits in the next life?
    3a) If so, why are *you* not convinced to believe in (and act to avoid) hell, given that many hell-fearing Christians (or Muslims) would suggest that *you* should apply the very same principle to *their* belief?

    5) What other “approaches to knowing” would you suggest are valid and reliable pathways to true beliefs, beyond the evidence- and logic-based processes of scientific and rational inquiry? Why is it anything more than special pleading to argue that we should adopt them for this particular question, if they are not at least as reliable as those methods?

    6) Is “I don’t know” a “satisfactory” answer (given that it’s true in most cases) to the questions of cosmology, biogenesis, etc.? If not, why not?
    6a) Is “god did it” substantially different to “I don’t know, but I know someone who does”?

    All the best, UnkleE

  48. Hi, well I didn’t think there’d be much agreement there. I’m totally amazed by your comment that ”the overall effects are negative” when so many studies say otherwise. So as I said, I think it is pointless discussing further, and I’ll consider that discussion has reached an end, and move onto the second matter I outlined. (I’m not addressing your questions directly yet, as I think responding to the three matters I outlined will cover all of them.)

    2. Should we adopt a “scientific” epistemology to determine whether we can believe in the existence of God?

    First, to clarify some apparent misunderstandings ….

    ”you want me to *lower* my standard of evidence below that I would apply to my choice of spouse, decisions for my healthcare, or even which car I should buy?!? “
    No, I don’t. I can’t see anywhere I have suggested that.

    ”Why do you think I should be willing to adopt a standard of evidence which is lower than the one you yourself apply to religious claims?”
    I don’t. I can’t see anywhere I have suggested that.

    ”Do you stand by your uncertainty-based principle of believing (or acting as if you believe) in order to reap certain benefits in the next life?”
    I don’t think that either. I think we must make decisions about God in the face of uncertainty, as we do in most parts of life, and I think the enormity of the claims of plausible religions gives an incentive to decide rather than keep on waiting for more evidence, but I certainly don’t believe in deciding against the evidence or acting/pretending.

    I think you have misunderstood what I have been saying. So I will try to summarise and see if we can conclude this topic also.

    1. I believe we should always act in accordance with the evidence as we see it. (There may be some trivial cases where we choose a “white lie” for social conformity sake, something I dislike. We can ignore this here.)

    2. I believe evidence can be a number of things – observation, experiment, authority of someone who has studied what I haven’t studied, introspection, personal experience, other people’s experience, intuition, maybe more. For a christian, revelation might be considered under authority or experience (personal or other people’s). These different types of evidence have different levels of (un)certainty in different situations and we should take that into account.

    For example, some questions are matters of fact, and on them the authority of experts is important, but authority is of little value on opinions or matters of fact outside the so-called authority’s area of expertise. Also, analytical thinking is most useful for resolving many questions, but if matters get too complex, psychologists have found that intuitive thinking can be better.

    3. In assessing evidence, we have to take various realities into account. How certain do we need to be? Are there risks, hazards, benefits, etc, associated with the decision that make it better to choose one or the other option if we lack certainty? Is there a time frame on making a decision?

    I’ll give a scientific example of this. I used to work as a river and catchment environmental manager. Most rivers in Australia are under environmental threat from agriculture and other land uses. With only a small population and a large land area, we can’t possibly collect for all rivers all the data we need to make fully science-based decisions on management. But rivers are deteriorating, and to wait for some nebulous future when we might have enough data would effectively destroy the ecosystems of many rivers. So we use appropriate decision-making systems, including these:

    (i) We use existing understanding to “triage” catchments, using fairly simple criteria to classify the environmental value and the social/economic value of each catchment. It becomes obvious in most cases which catchments are more critical, and we start with them. Even though we know there’ll be some catchments we’ll rate wrongly, we’ll mostly get it right.

    (ii) Instead of waiting for statistically valid scientific data, we use rapid assessment – we employ “expert panels” of experienced water scientists to visit the catchments that are highest on our priorities developed in (i), and make fairly subjective assessments of various aspects of river health. This allows us to gain a fair picture of crucial issues in the critical catchments.

    (iii) We undertake adaptive environmental management – we begin management actions plus collection of better data in the critical catchments and make ongoing assessment of whether the critical rivers are responding to management in the way we would predict and hope. And then adjust management on that basis.

    This approach parallels a fully scientific approach, but does it on the cheap and in far less time. It generally makes the best decision, but will occasionally spend money in less critical catchments, or fail a catchment that wasn’t properly identified as critical. As you can see, it uses provisional data, authority, experience and even intuition rather than wait for complete scientific evidence.

    4. I suggest we use similar processes, less formally defined, in most of life. We “triage” political issues when we vote, we select life partners without having anywhere near a full knowledge base, we make decisions (e.g. about careers) as best we can and then adjust our plans as we gain experience. And so on.

    5. I am suggesting exactly the same with God. i.e. We can’t consider every possible philosophy and religion, so we consider the ones that seem most plausible, and adjust as we go. We don’t ignore any evidence as you seem to think I am suggesting, but we don’t wait around our whole life for new evidence that may not eventuate. Rather, we decide on the basis of the evidence we have what looks most probable, examine it further, and make a choice if we believe it is more than 50% probable and more probable than any other option. We can always change our mind.

    6. We use all the evidence and all different types of evidence, weighing each one appropriately according to its uncertainty. For example, we use the best scientific and historical experts (authority) to understand important facts. We accept ours and others’ experience as evidence, with suitable open-minded mild scepticism. e.g. there are hundreds of millions of miracle claims. Many could possibly be rejected if we had time to assess them, but we don’t. So we simply assess some of the most plausible and best documented, and ask whether they would be more likely to occur if God existed than if he didn’t. We don’t necessarily reject them because of uncertainty, because even uncertain events can be a pointer to truth if there are enough of them.

    I’ll say again, none of this is to ignore any of the evidence (that is actually what sceptics often do) but to consider it all with appropriate probability.

    I suggest this is an appropriately scientific method for this question. I wouldn’t use the same method for testing thalidomide or writing a textbook, but I would use it in other situations. And I suggest most people would too.

    Is there anything there that you can agree with?

  49. Hi UnkleE,

    When I say that the overall effects of religion are negative, I mean that in aggregate. If religion offers some benefits (such as increased prosocial behaviour), but only does so as well as do equivalent secular organisations (which the best research is entirely consistent with), then, based only on considering the benefits, religion is *equivalent to* non-religion. If, however, religion also introduces new harms that are not present with non-religion (Islamic terrorism, abortion clinic bombings, children dying at the hands of faith healers, and so on), then it is *worse* than non-religion. It goes back to the discussion we’ve had on risk-benefit analysis – if there are no benefits beyond placebo (if it’s not better than secular organisations), but it does additional harm, then it’s a negative intervention.

    Based on that principle, I would even hold that, to a certain extent, religion could offer benefits above and beyond non-religious organisations (which I do not think the evidence supports) and still be worse – how much pro-social behaviour does a religion have to encourage before it balances out the cost of lunatics blowing up one doctor, let alone the existence of ISIS? You can’t rationally look at the benefits in isolation from the harms, and small benefits do not justify significant harms (and if there are no specific benefits above and beyond the “placebo” – in this case, non-religious organisations – then *any* harm, no matter how small, is reason enough to discard the whole thing).

    But, moving on…

    Your response to 1 and 1a seem to ignore some of your own comments. You’ve talked about “other ways of knowing” (without defining that, or explaining how they are equally or more reliable methods for reaching true conclusions than scientific rationalism). You’ve criticised my “over rationalistic approach” and claimed, without evidence, that I don’t seek to apply it to other aspects of my life – thereby implying that I shouldn’t apply such a high standard to the most important questions we can ask, simply because I’m imperfect at doing so for other questions. And you’ve failed to acknowledge my repeated comments on Loftus’ outsider test for faith. That alone speaks to encouraging me to accept a lower standard for belief than you do – because, as I’ve noted, I’m reasonably confident that you would apply at least that standard to claims made by, say, a Hindu.

    Let me go through this point by point – I agree with much, but we disagree on many points too.

    1) Yes, we should absolutely behave in accordance with the best available evidence. I would add the caveat that we should be very careful in deciding what constitutes evidence, how we assess and interpret the evidence, and how much confidence we place in any conclusions we draw.

    2) What constitutes evidence will absolutely vary from one situation to the next, as will the strength of each. However, when it comes to matters external to our own minds, a thing should only count as evidence if it is likely to be materially relevant to *any reasonable person who might consider it*. You say that “[f]or a christian, revelation might be considered” – but I would argue that you shot yourself in the foot the moment you had to qualify who would consider it evidence. Obviously there’s going to be disagreement amongst people – you and I might see the same object but observe different features, and disagree on it later, but at least in principle there’s something we can go back to in order to adjudicate the dispute (even if we can’t do so in practice) – but a reasonable person with access to the same information should be able to at least say “yup, I can easily see why you find that convincing” (even if they disagree with your conclusion).

    I would also strongly disagree that intuition constitutes evidence. Intuition is a process by which we reach conclusions, it is not evidence for those conclusions. Intuition *can* be a reliable method for reaching conclusions, primarily in the mind of an expert. For example, a doctor might make an accurate diagnosis without conscious awareness of the facts on which he based his conclusion – but his intuitive diagnosis is not evidence that his diagnosis is correct! I have little issue with the other types of evidence you mentioned, so long as their limitations are understood and the information they offer is treated appropriately.

    3) I completely agree with your principle. I would quibble with your description of the approach as “paralleling fully scientific”, however – triage of this type is entirely scientific. You have a hypothesis (sensors in these handful of areas will provide an adequately accurate model of the entire system), you test it out and you refine your hypothesis (adding, moving or removing sensors) until you settle on a model that works to the required level. Science isn’t about perfect knowledge – it’s about developing and refining models of reality in order to best understand and explain the world around us. Sometimes we want highly accurate models; other times we want models that are reliable enough to serve a purpose. Both can be entirely scientific.

    4) Agreed.

    5) This is, of course, where we seriously disagree. When we’re talking about the health of rivers, or the benefits of drugs, or the emotions of other people, we’re talking about things for which we have *strong* evidence of their existence (effectively undeniable, limited only by problems like solipsism), and therefore we can justify using limited models to assess their status for practical purposes (so long as we recognise and acknowledge the limitations of our models, of course, and don’t place undue confidence in them). We also have evidence about the risks associated with *not* behaving in that way. But for god, we have *none* of those things – we have no good evidence to show he exists (or is even possible), and we have no good evidence to show the consequences of not behaving as if he does. Again, I’ve spent a lot of time on this subject, and I’m yet to come across a single piece of evidence that passes even the very lowest bar of Loftus’ test (let alone any of the bars you’d need to cross in order to justify belief in the supernatural, *let alone* the subsequent bars to get to justified belief in god!).

    Furthermore, you’re comparing situations in which we *need* (to whatever extent and for whatever reasons) to make a decision, with a situation in which there is no obvious need to decide. We cannot solve the problem of hard solipsism, for example, but we need to act as if it is false in order to deal with the world as it appears to us (otherwise we seem to get hit by a bus we don’t believe in). We know that our models are limited, but we care about the health of our rivers and so we use those limited models as best we can. We can’t be absolutely sure that our partners love us, but it’s probably more destructive to the relationship we want if we ignore the evidence that they do rather than focus fearing the possibility that they don’t. *But* we do not need to believe in god in order to deal with the world as it presents itself to us – we can without question live entirely productive, ethical and meaningful lives as atheists. We also don’t need to believe in order to gain any benefits in “this world” – despite our disagreement on the benefits of religion, I think we can both agree that it’s at least possible to get the benefits without the beliefs; and we know that belief can have severe negative consequences in “this world” – and based on that, I think it entirely reasonable to conclude that, considering only “this world”, we cannot justify even belief for practical purposes. In order to justify your principle, then, you need to appeal to benefits or consequences in the “next world” – but you cannot do that. There’s no demonstration that a “next world” is even *possible*, let alone that one in fact exists, what it might be like, what the rules are, how you get there, and so on, and appealing to an unknown when you don’t have an obvious and immediate need to decide is entirely irrational. You cannot justify belief in an unknown on the basis of a bunch of other unknowns – because I could simply say “well, there are *heaps* of unknown reasons why *not* believing could be a better path to the afterlife – perhaps god reveres and rewards people who utilise the rationality he instilled them with above those who believe things in the absence of evidence”! Just as I’m sure you’re not convinced to give up your belief by that claim, I’m not at all convinced by your appeals to the unknown – my point is that *you* shouldn’t be, either, because your claims about the afterlife are equally absurd!

    Also, I must continue to emphasise that there’s no probability attached to existential claims – something *is or is not*, there is no “half-existing”. We can talk about our confidence that a thing exists, but not the probability that it does. I’m highly confident the earth is a spheroid; I’m not at all confident that it’s snowing in Moscow right now, and I’m reasonably confident that it’s not going to rain where I am in the next hour – but all of those things are or are not true! Probability only applies to existential claims about the future (whether something will *come to exist*), not about the past or present. But back to the point, if you have greater than 50% confidence in a claim then it *must* be the most convincing (to you) explanation. I assume you’re thinking of comparing Muslim god with Christian god with atheism with…, but really, you should be first considering “is there a god?”, and, if and only if the answer is yes would you then ask, “which god is it?”. You need greater than 50% confidence for both of those claims (for example, I’m at 0% for the first, and therefore the second is meaningless to me).

    6) I agree that we accept and weight evidence on various matters, as discussed at (2). I agree that we accept claims of personal experience as evidence of various things. I *disagree* that miracle claims constitute evidence for the existence of the supernatural, let alone for god, and I also disagree with your description of how we handle them.

    A miracle claim is evidence that Fred believes he observed a miracle – but it’s not proof of even that (Fred could be lying about his belief). But taking his word, all we know is that Fred *thinks* he saw a miracle – we have no evidence that the event he described actually occurred (perhaps it was a hallucination, a dream, or he was mistaken), let alone that it was a miracle. So, when we have a miracle claim, the first thing we need is to establish that the event actually occurred as described. Sometimes that’s not possible – and we should *immediately* reject those entirely unverifiable claims as worthless to the question of god’s existence (because if we can’t even confirm that an event occurred, there’s no point in looking for a cause). After this most basic triage, we would then need to examine the event in detail, and compare other accounts with Fred’s story, looking for simple problems (like common cognitive failings, misunderstandings, and so on) that would explain why Fred *thought* it might have been miraculous (and obvious and *common* example would be Fred claiming it’s a miracle because he prayed then his cancer was gone, ignoring that he’d been taking a cocktail of chemotherapeutics which we can reasonably demonstrate to cure at least some cancers with or without prayer). We might also look for the *actual* cause of the event (not just why Fred might have been misled), if it’s possible to do so. And once we rule out *all* known explanations, we stop – and we say “I don’t know why this happened”. We *do not* go further and say “well, I can’t explain it so it must have been god/aliens/pixies/interdimensional time travelers”! We might *propose* those as hypotheses for further testing, but we do not *claim to know things that we don’t know* (at least, not if we’re being intellectually honest).

    It doesn’t matter how many miracle claims you stack up, it doesn’t matter how good the evidence that they happened, and it doesn’t matter how inept we are at finding the causes of the events in question – the fact that we don’t know what caused them is not evidence in support of any claim about what caused them! I can take every miracle claim that’s ever been made and I still have *no* evidence for god’s existence – all I have is a bunch of evidence that lots of people *think* god did something, and that’s not really of interest to either of us, I hope. Every verifiable one would need to be investigated and *all* known possible causes somewhat excluded before they would even be an interesting nudge toward considering *unknown natural causes*! You’d need to *definitively* rule out all natural causes – known and unknown – before even considering a supernatural cause, let alone a god, *let alone* the Christian god!

    In addition, when it comes to a claim like god, we most certainly do *not* “ask whether [an event] would be more likely to occur if God existed than if he didn’t” – because that’s a meaningless question. We live in a universe where god does or does not exist. Let’s assume that some event, X, that Fred thinks is miraculous, occurred. If god exists, then it is 100% likely that X occurred – because it did. If god does not exist, then it’s 100% likely that X occurred – because it did! The fact that an event occurred does not constitute evidence for its cause – if we want an event to constitute evidence for a hypothesis like “god did it”, then we need to *predict* the event, instead of rationalising one we can’t explain post hoc. And (again) probability applies only to future events (probability is *predictive*, not *descriptive*, of the events it considers). So, we *could* say “if god exists, then praying will help me recover from surgery faster” – and we can test that (and have done, and the evidence does not support the claim), but we *cannot* say “I got better as quickly as I did because I prayed” (because how can you rule out having a mild case, having a strong immune response, being particularly responsive to the treatment, having been misdiagnosed, or any number of other possible causes for your apparent quick recovery?).

    Post hoc rationalisation is a perfectly good way to *generate* hypotheses for further testing, but it is *not* evidence in support of your hypothesis. If you want to say “prayer makes people heal faster”, that’s great – we can observe people praying and then getting better, sometimes faster than people who don’t pray, so it’s a reasonable hypothesis. But the *fact* that people pray and get better is not *evidence* for your hypothesis – you need to *test* your hypothesis in order to generate that evidence!

    Also, I’m certainly not meaning to imply that you’ve suggested we ignore evidence – in fact, I agree with you that many skeptics (myself included) think we should ignore much of what *you* might consider evidence (due to a lack of verifiability or reliability, for example). I think all evidence needs to be weighted – but we weight it on the basis of known facts about reality, not in light of the hypothesis we’re trying to prove. I think your bar for what constitutes evidence is *far* too low when it comes to the existence of god – you accept claims that have little or no evidence to support them (e.g. that a “next life” is possible, that the physical constants in our universe *could* differ, and indeed that a god could – let alone actually – exists), and I think that you wouldn’t apply such a low bar to other aspects of your life, *even if* you apply the principles you’ve laid out here. You wouldn’t (I hope) accept homeopathy as a valid treatment for disease simply because lots of people say it works, nobody has proven that it doesn’t, there’s lots of published literature favouring it, the guy who originally developed it seemed like a genuine sort of person, it’s been around for a long time, and your cold got better after downing some oscillococcinum – but you accept *precisely* that type of low quality evidence in order to justify your much more extraordinary belief in god! I think you’re being inconsistent in applying your principle, watering it down for the most important questions where the more reasonable approach would be to crank your skepticism to 11!

    I look forward to your next response, UnkleE – although I would note you’ve now only covered questions 1 (but not 1a or 1b), 3 (but not 3a), and 4 of my original list. We’ve touched on 2 (but not 2a), although only to the extent that you’ve denied saying or implying that – which I continue to dispute.

    All the best.

  50. Hello again,

    Looking at your response to my second area of discussion – epistemology – it is good that we can at least agree on a few of those points. But of course there is no agreement on the most important ones (for this discussion at least). Which means that before we even get to the question of God, we are in disagreement about how to even consider this.

    Again, I think your approach is not truly evidence based. You introduce, without any reasonable justification that I can see, a different epistemology for God on the unjustified grounds that we have “no good evidence to show that he exists”. There are two problems with this approach. (1) There are many things that we have no clear evidence whether they exist, but we still address them just the same, and (2) you have assumed the result (no evidence) before you start, by making that statement part of your epistemology. So it is a circular, self-fulfilling epistemology.

    So having said that, I propose to leave epistemology behind as I have left behind fine-tuning and the social benefits of religion, considering them to be topics we will not be able to agree on while you (and I) hold the epistemologies that we each do.

    So onto my final matter …

    3. Do either of us make assumptions or start with a preference for belief or unbelief?

    Now clearly we both do, everyone does. No-one is 100% open-minded, we all initially defend our existing view, even if we later change it. So the question is (I think) do either of us unduly do that?

    I will admit that while I try to consider evidence fairly, and have changed my views on many important matters during my life as a christian, I’m sure I colour my thinking by loyalty to God/Jesus. I think that is consistent with my view that my belief is not just an intellectual thing, but also a relationship with a personal God. It’s not something I talk about that much on this blog, because it isn’t relevant to a discussion of facts and evidence, bit I’m sure it’s there. And I continue to hold that view about God being personal because I think the evidence confirms it.

    So what about you?

    To sum up, I have found your views to be far less evidence-based than you have claimed.

    (1) You continue to hold views about the science of fine-tuning that almost all cosmologists disagree with.

    (2) You continue to say religion is detrimental even though almost all the experts say that it isn’t. You argue that all the benefits of religion can be obtained by non-religious people even though the experts say (i) some extra benefits accrue to sincere believers, and (ii) those theoretically possible benefits aren’t experienced by so many non-believers in actual fact. Even your latest comment shows a lack of evidence. You mention “the cost of lunatics blowing up one doctor, let alone the existence of ISIS” seemingly unaware that most experts have concluded that most wars and most terrorism is NOT primarily religiously motivated but motivated by other grievances. (Having said that, I will agree that religion can do harm – just not the sort of religion I am talking about!)

    (3) Then you use a circular epistemology that biases your conclusions by treating the question of God differently to other questions on the basis of evidence (or lack of it) before you have even looked at the evidence! (Because epistemology is about how we approach evidence.)

    So it seems to me that you are actually less evidence-based than I am, even granted my pro-God bias. You ask me why you would do such a thing, but I am not a psychologist and I don’t know you, so how could I answer that? But I can say this:

    (a) It looks like that is what you are doing.
    (b) When I was a teen without a christian upbringing deciding what I believed, I was very resistant to allow a God to interfere in my life.
    (c) Atheists like Thomas Nagel and former internet friend Steve, have been quite clear that they don’t WANT God to be there.

    So I can’t say any more than that.

    So I think it may be time to end this discussion if that is OK with you. I have appreciated the opportunity to discuss, and your ongoing courtesy. But I really can’t see any hope of us agreeing anything important granted the problems I have raised here (and I imagine your disagreement with my observations!).

    So shall we call it quits?

    Best wishes.

  51. Hi UnkleE,

    Let me first point out that the *beginning* of our discussion was predicated on a disagreement over epistemology. I put forth that you had a burden of proof that atheists lack – everything after that was a degeneration from that topic.

    Turning to your criticism of my epistemology. I agree that there are many things for which we have no concrete evidence but which we assume to exist – but only when we need to actually do so. Again – nobody has disproved hard solipsism, but we *need* to respond to the world as it presents itself, even if we’re responding to an illusion. But we don’t generally accept existential claims when there’s no pressing need to do so – we don’t believe in unicorns because there’s no apparent need to accept their existence in order to live our lives. God is the same – I can respond to reality as it presents itself without any need to posit – let alone accept – a god, and so in the absence of convincing evidence that god exists, I do just that.

    I’d *definitely* like to argue against your accusation that my epistemology is circular. I have in *no way* assumed the result (“no evidence”) prior to enquiring on the subject. Of course, since that enquiring started long before I stumbled upon your blog, you wouldn’t know that, but that doesn’t mean you should assume that I’ve not asked the questions. My initial stance, having lived in practice (if not in name) as an atheist my entire life, was “*I* have no evidence for the existence of god – let’s see what others bring to the table”. And, despite much reading, viewing discussions and debates, conversing with believers of a handful of different faiths, and tertiary studies in philosophy of religion, I’ve found not a single thing that crosses Loftus’ outsider test, let alone anything actually convincing. Given this, I therefore *do not believe* that a god exists – whilst remaining entirely open to the possibility, and *entirely* neutral with respect to a non-intervening god (admittedly I’m less neutral towards, say, the Christian god due to the existence of *dispositive* evidence – e.g. problem of evil, ineffectiveness of prayer – but I’m still open to the possibility that even he might exist). Again, “atheist” means “I am not convinced” – it does *not* mean “I am convinced of the opposite”.

    In addition, I’d like to point out that *many* atheists, myself included, would be entirely pleased to find out that there’s a god, an afterlife, and so on. Now, I’d admittedly be rather unhappy if it turned out to be the jealous, vindictive, evil being described in the Bible or the Quran (although I would still *believe* in him if the evidence were sufficient – I’d just refuse to worship him), but if it were the god that the average Christian believes in (minus any punishment of non-believers, which would be an act of outright evil), that’d be perfectly great. Do you think atheists, for the most part, *want* to cease existing? Certainly, as you point out, there are examples – but there are plenty of counterexamples, too, so I certainly don’t think you can make any type of blanket statement. I’m also not resistant to god “meddling” in my life – he either does or he does not, and my knowledge or belief on the matter changes that fact not one bit (and since I’m reasonably convinced that humans lack meaningful free will anyway, I again fail to see how it makes a difference to me whether god or physics determines my life). I’d also challenge you to explain how it “looks like” I’m rejecting god on principle, when *every* post I’ve made has focused on evidence and justification for belief. I do not believe in god *because the evidence is lacking*; I do not believe the evidence is lacking because I am committed to disbelieving in god. To use a common turn of phrase, I’m married to the process of rational inquiry, not the conclusion. The process, whilst imperfect, leads more-or-less reliably to valid conclusions, but it also allows us to update them when new information comes to light – but if we strongly commit to a conclusion, then it’s easy to warp the process (intentionally or not) in order to maintain a belief.

    My epistemological standard is really simple. If there is objective evidence for a claim, I will accept the claim to the degree that is justified by that evidence. That degree will depend primarily upon the reliability of the evidence, the evidence for alternative claims (or alternative explanations for the evidence in question), and the extraordinariness of the claim. I will, where required, act in the face of a severe lack of evidence (although I try to avoid acting *against* the best available evidence) if I cannot reasonably wait for better evidence to be available. Since there’s no obvious pressing need to act as if god exists, I’m not going to do so simply because people hove posited a bewildering array of benefits and consequences for that decision – because propositions are not evidence!

    Let’s get to your final point. Of course we come at an argument from our own starting positions – otherwise we wouldn’t be arguing. That said, which of our worldviews is likely to skew the perception we might hold more? For you, you’ve got a reasonably firm conclusion in mind – there is a specific god, who does X, wants Y, and so on. You therefore have an incentive (no matter how unconscious and small) to view facts in light of that conclusion – you’re predisposed, for example, to concluding that prayer works because you believe that a god who answers prayers in fact exists, so prayer working would be entirely consistent with your worldview. In contrast, I’m *not* committed to a conclusion (beyond the simple emotional concern about being wrong – but, as a research scientist, I find I’m wrong about things I’m an actual PhD-qualified expert in more days than not, so I don’t see that concern as having a *huge* impact). If prayer works, then it works and I would accept that. I am entirely content to follow the evidence wherever it leads – and if it gets me to god, that’s great, but I’m *not* going to treat the question with any less respect than I would when considering a medical treatment or a used car. I’ve changed course on strongly held beliefs in the past – and my position on the existence of god isn’t even a belief!

    To your criticisms of me personally:

    1) I *do not* hold views which run counter to the consensus view of cosmologists (to the best of my knowledge, anyway, given that I’m effectively limited to “popular” cosmology). I *do* argue that extrapolation from facts about our universe to hypotheses about what *might be* beyond our universe does not provide justification for claims about what *is* outside our universe (if there is such a place). The very moment someone explains to me how cosmologists have *tested* the claims that they make about what exists outside our universe (that is the *actual* claims, not the evidence from our own universe which supports them), I’ll accept them; until those tests are developed and executed, all cosmologists have is a lot of evidence about our universe which *points toward* things outside it, but no evidence to actually distinguish between the many different and mutually exclusive hypotheses about what is or is not “out there”.

    As I tried repeatedly to explain, I fully accept the evidence-based conclusions that cosmologists report about *this* universe and about what that *may* mean for what’s outside it, but I also recognise the difference between a *fact* and a *hypothesis* – a distinction you’ve repeatedly failed to acknowledge in this context. Going back to our discussion of the cosmological argument, it’s a *fact* that the universe is compatible with life; it’s a *hypothesis* that this compatibility is the result of fine tuning (or chance, or a multiverse, or…). Until we have evidence in support of that hypothesis (i.e. *actual* evidence of fine tuning beyond the simple fact of compatibility with life), we are not justified in claiming the universe is fine tuned. You’re still attacking a straw man of my position, which I find disappointing after so much good-faith discussion.

    2) You don’t get to cherry pick “the sort of religion [you are] talking about” from “religion” as a whole. Every comment I’ve made in this entire discussion, except where I’ve otherwise specified, has been about *religion*, not *your* or even *a* religion. If we’re talking about the benefits of religion, we’d best be talking about the benefits of *religion in general*, because to do otherwise is nothing more than special pleading given how broadly people interpret that word. I also never argued against (i) or (ii) as facts – although I certainly *did* (and do) object to your presumed reason for (ii), favouring the evidence that social factors, rather than religious ones, are responsible for all of the benefits that have been adequately studied – and likely for those that haven’t. But turning to ISIS and abortion clinic bombers – it doesn’t necessarily matter what the leaders of movements are inspired by, it’s a simple fact that at least some of these people claim adamantly that they’re doing whichever god’s work, and unless you’re going to tell me you have special insights into the minds of these people, I’m afraid you can’t reject the evidence that indicates at least some of the harm done by these individuals is the result of their religious beliefs. Again, I only need one religiously-inspired killing to counterbalance a *lot* of social benefits (which, based on the balance of evidence available, I do not think are specific to religion). So, again, you’re presenting a straw man of my position (I never said there were no benefits, or that they were not frequently associated with religion – just that the “religion” part is superfluous), along with cherry picking (focusing on *your* religion, when most believers out there would be far more adamant than I that you’re wrong – some might even kill you for it) and special pleading (discounting individual motivations for action simply because, say, at the higher levels wars aren’t primarily inspired by religion).

    3) As discussed, my epistemology is in no way circular. I have no firm conclusion to “does god exist?” (I am entirely agnostic), but I also see no reason to accept the claim (and am therefore an atheist). I don’t *choose* to be an atheist (because we do not choose our beliefs) but I *am* one by definition. My lack of belief stems from the fact that the evidence is unconvincing – I do *not* find it unconvincing because I lack belief. If you really want to at least slightly test that, I’ve asked you many times now – identify a single piece of evidence which passes Loftus’ outsider test. My *practical* conclusion (I will act as if no god exists) is based purely on the fact that I’ve found no evidence that remotely justifies rejection of the null hypothesis, and until I find something that does I’m not going to sacrifice any part of the perfectly satisfying life I *know* I’ve got chasing one of the many that various self-proclaimed prophets, mystics and plain old drug-addled hippies have promised me.

    The only way I can possibly understand your criticising my epistemology as circular is if, despite all our discussion, you’re still convinced that I believe “no god exists” – which is simply not the case, as I’ve repeatedly pointed out. I do not understand how the philosophical positions “I do not know if a god exists” and “I am not convinced that a god exists”, combined with a practical attitude of favouring the null hypothesis in the absence of dispositive evidence *and* a repeated focus on basing conclusions on evidence, could lead me to being committed to disbelief. I engage in discussions like this to *test my own knowledge and assumptions* and to perhaps learn something new, not because I want to promote my own viewpoint (there are *much* better ways of doing that than commenting on a random blog post!).

    So, that brings me to the end of your post. Let me put it to you that you’ve still skipped over most of my recent questions, specifically:

    1a) What evidence would get you to the point of complete uncertainty (i.e. 50:50 on the existence of god)?
    1b) What would convince you to discard your beliefs as unfounded (without necessarily rejecting the possibility that god might exist – which is the position that I am in, as an agnostic atheist)?

    2) Why do you think I should be willing to adopt a standard of evidence which is lower than the one you yourself apply to religious claims?
    2b) Is there a single piece of evidence that you can offer which passes the outsider test for faith?

    3a) … why are *you* not convinced to believe in (and act to avoid) hell, given that many hell-fearing Christians (or Muslims) would suggest that *you* should apply the very same principle to *their* belief?

    5) What other “approaches to knowing” would you suggest are valid and reliable pathways to true beliefs, beyond the evidence- and logic-based processes of scientific and rational inquiry? Why is it anything more than special pleading to argue that we should adopt them for this particular question, if they are not at least as reliable as those methods?

    6) Is “I don’t know” a “satisfactory” answer (given that it’s true in most cases) to the questions of cosmology, biogenesis, etc.? If not, why not?
    6a) Is “god did it” substantially different to “I don’t know, but I know someone who does”?

    I would add one more to the list, in light of your most recent post: what do *you* think that *I* believe?

    If you’d care to discuss those, or respond to any other part of this post, I’d be more than happy to continue the discussion – if nothing else, I’d love to hear how you get around my objections to your own epistemological standards (raised in response to your previous post on “Should we adopt a “scientific” epistemology to determine whether we can believe in the existence of God?”). However, if you’ve no further interest, I’d like to say that I’ve enjoyed talking with you, and appreciated the almost entirely civil tone we’ve managed to maintain. I look forward to any response, but if it’s simply “bye”, then I’ll simply say thanks for the constructive (if unproductive) debate, and wish you all the very best.

  52. Hello again,

    I think our respective responses show that we are talking past each other and little is being gained, I’m sorry to say. So here I will sum up and try to draw to a close. I will briefly answer your questions because doubtless you will feel I have avoided them otherwise, but I will be brief because I think I have answered them already.

    1a) What evidence would get you to the point of complete uncertainty (i.e. 50:50 on the existence of god)?
    1b) What would convince you to discard your beliefs as unfounded (without necessarily rejecting the possibility that god might exist – which is the position that I am in, as an agnostic atheist)?

    (a) If the universe didn’t exist, if it wasn’t fine-tuned, if the human race had no rationality, free will, consciousness, sense of true ethics, if some people weren’t healed when they prayed, if I didn’t experience God, if Jesus never lived or was not a good man.
    (b) Ditto.

    2a) Why do you think I should be willing to adopt a standard of evidence which is lower than the one you yourself apply to religious claims?
    2b) Is there a single piece of evidence that you can offer which passes the outsider test for faith?

    (a) I don’t. I have said that before. I think you are the one with inconsistent epistemology (see answer to 5).
    (b) All of the above.

    3a) … why are *you* not convinced to believe in (and act to avoid) hell, given that many hell-fearing Christians (or Muslims) would suggest that *you* should apply the very same principle to *their* belief?

    Because I don’t believe it’s true, but I believe Jesus is true.

    5) What other “approaches to knowing” would you suggest are valid and reliable pathways to true beliefs, beyond the evidence- and logic-based processes of scientific and rational inquiry? Why is it anything more than special pleading to argue that we should adopt them for this particular question, if they are not at least as reliable as those methods?

    Authority, experience, introspection, intuition are all helpful. These are all evidence-based to a degree, but I still rely most on what people call evidence. It isn’t special pleading because we use these other methods in all of life when they are appropriate, so why not allow them when considering God?

    6a) Is “I don’t know” a “satisfactory” answer (given that it’s true in most cases) to the questions of cosmology, biogenesis, etc.? If not, why not?
    6b) Is “god did it” substantially different to “I don’t know, but I know someone who does”?

    (a) Of course. But it isn’t the best answer when there is a hypothesis that does provide an answer. This is basic science and logic.
    (b) Yes, if a logical argument can be constructed – which in the examples I reference can be done. “I don’t know” without a refutation of the argument is just avoiding the conclusion.

    I would add one more to the list, in light of your most recent post: what do *you* think that *I* believe?

    I don’t presume to know what you believe apart from what you say in your comments. Several times you have suggested I had assumed something when in fact my comments were based on what you said (and I generally quoted from you).

    I’m sure you won’t be satisfied with those brief answers, but I have said all those things in more detail before, so I think that is enough.

    So now to sum up.

    I still feel that your approach makes it almost certain that you’ll come to the conclusions that you do.

    1. My observation about fine-tuning has nothing to do with the conclusions we each draw, or what is “out there”, but from your non acceptance of the science. You say ”Until we have evidence in support of that hypothesis (i.e. *actual* evidence of fine tuning beyond the simple fact of compatibility with life), we are not justified in claiming the universe is fine tuned.” whereas Lewis & Barnes say: ”With every step forward in science, these fine-tuning issues have become more significant. We find ourselves questioning the nature of many of the things we take for granted, from the fabric of space and time, to the mathematical underpinnings of the Universe. At every level, we find that our Universe’s ability to create and sustain life forms is rare and remarkable.” I suspect you still don’t understand that this fine-tuning is a scientific thing, not an assumption about cause, even though I have said that over and over.

    2. Likewise on the benefits of religion, you offer all sorts of comments to show there can be disbenefits and other non-religious ways to achieve some of the benefits, which I agree with. But still you ignore or refuse to accept what Lilienfeld & Ammirat, Connoir Wood, Andy Newberg and others have concluded.

    3. Finally, you are not happy that I have “assumed” your epistemology is circular. But I based the comment on your statement, discussing epistemology and why you require a higher standard of evidence for God than you do in many other parts of life, that ”we have no good evidence to show he exists”. Thus you justify a critical epistemology by already begging the question of whether there is any evidence. You don’t have evidence (perhaps) because of your epistemology, and you justify your epistemology because we don’t have evidence. That seems to be circular by your own statements.

    These observations, to which you will have responses, encapsulate our differences, which I think are irreconcilable. So I really don’t think there is a lot of point in continuing this discussion. I don’t think we are going anywhere, the comments are very long, and responding takes time. I don’t want to suggest any blame or fault in this, and as I have said before, I really appreciate how courteous this discussion has been. But I think it is time to call a halt.

    So I am happy to read any final response you make, but I don’t propose to follow up any further. I hope that is OK with you. Thanks.

  53. Hi UnkleE,

    Thanks for the response. I must say, however, that I’m rather frustrated by your reply – after such an interesting conversation (albeit one that went down many rabbit holes), where I think we’ve both managed to remain largely friendly, you’ve now come out with some incredibly hard to swallow responses.

    1a/b) After all the accusations about my own closed-mindedness, you put forth that *nothing* could make you change your mind (so long as your mind – and thus whatever universe it exists in – exists, you will believe in god). I’m sorry, but I feel compelled to point out the utter hypocrisy of that.

    2a) Allow me to copy-paste my previous response: You’ve talked about “other ways of knowing” (without defining that, or explaining how they are equally or more reliable methods for reaching true conclusions than scientific rationalism). You’ve criticised my “over rationalistic approach” and claimed, without evidence, that I don’t seek to apply it to other aspects of my life – thereby implying that I shouldn’t apply such a high standard to the most important questions we can ask, simply because I’m imperfect at doing so for other questions. If those things are not asking that I accept a lower standard of evidence, then I don’t know what is.
    2b) I’m sorry, but I continue to disagree, for the simple reason that I’ve had Muslims, Mormons, Christians, Hindus and Buddhists offer up the same (or substantially similar) evidence to justify belief in *their* (mutually contradictory) fairy tales. That’s the very *definition* of failing the outsider test. The point of the outsider test is to say to *yourself*, “if a believer of a different faith told me that this evidence proves their god exists, would I find it compelling?” If not, the evidence fails the test – and it is the most *generous* test we can apply to *any* claim, let alone the extraordinary claims of religions! If you’ve never listened, allow me to recommend you find The Atheist Experience online show ( – I listen to it as a podcast), and listen to it for a few weeks (episode 21.10 from March 12th – find a link at – was particularly interesting on this very topic, although the outsider test was not discussed by name) to try and understand why outsiders find the evidence that you or other believers point at to be entirely unconvincing.

    3a) Again, you’ve accused *me* of circular reasoning and flawed epistemology, but your own boils down to “I don’t believe the other things because they conflict with what I *do* believe”?

    5) As I said previously, intuition is a process for forming beliefs – it is not evidence for them. The others can be evidence – although they’re generally the *worst* types of evidence (with the exception of introspection as applied to our own internal mental states, for which purpose it’s the most reliable type). But these are not “other ways of knowing” – they’re types of evidence that can contribute to a scientific understanding of the world. What we *do* with the evidence they provide us matters – if we’re granting them a high degree of confidence (which is usually not warranted by these types of evidence) simply because they tend to give the desired answer (when other, more objective types of evidence – like reproducible experiments – do not), then *that* is special pleading and cherry picking.

    6a) “I don’t know” is *absolutely* the best answer when we don’t know something, whether we have a hypothesis about it or not! During my studies, I produced a number of drug-like compounds and, based on some primary pharmacological data, I had a *strong* hypothesis that they would activate a certain receptor. If asked, should I therefore have said, “these drugs will active receptor X”? Or should I have said “I think they may, but I don’t know”? ‘Cause I tell you what, my hypothesis was as wrong as it could have been! *Why* would any rational person accept an untested hypothesis simply because it hasn’t been proved false, or because they cannot think of an alternative? More relevant to religious claims, your hypothesis is far from the only one (it’s just the one *you* find most convincing) – why should any rational person favour one untested hypothesis over thousands of others? Why should we accept *your* hypothesis over the Muslim hypothesis, or the hypothesis that there is an as-yet-undiscovered physical law that dictates the types of universes which can exist?
    6b) I have no idea what you mean by this, I’m afraid.

    7) Despite your objections, by engaging in a conversation about belief you *must* assume what the other person believes. Hopefully, you base your assumptions on what they tell you – but you cannot engage with a position without attempting to understand it, and since it’s unlikely that anyone can perfectly understand a position held by another, you have to make some assumptions to fill the gaps. However, what your comments suggest is that *you* believe that *I* am committed to disbelief – that is, that believing that there is no god is part of my worldview – which is simply not the case. I would reiterate, I do *not* believe that there is a god, and I do *not* believe that there is no god – just as I do not believe that the number of jelly beans in that jar we found is odd, and I do not believe that the number is even. I am an atheist by *definition*, not by choice or desire.

    And to your other responses, I’m again thoroughly disappointed by the turn you’ve taken.

    1) Again, evidence trumps the small minds of humans. We can think and say it’s fine tuned all we want. We can take all the evidence we have about how unlikely the universe might be. But none of that can give us any confidence about what goes on *outside* the universe – a place, if it exists, that we cannot currently investigate – nor does it tell us anything about the *cause* of our universe. I’m not denying science, or the opinions of scientists – I’m arguing that many inferences from current science (and *all* of the inferences that would be required to get you to “fine tuning”, let alone “God”, are epistemologically invalid). We know physics is woefully incomplete – we may yet discover facts which give us access to things outside or before our universe, and when we do I’ll be all over them (as a lover of popular cosmology), but until then all we have is physicists taking facts about our universe and weaving stories about others – which is epistemogically useless. Again, you accuse me of allowing my belief to determine my understanding – but all that I’m doing is saying “I don’t know”, instead of assuming that some hypothesis (the universe is fine tuned) must be the case simply because it comports with my ideology.

    2) As I thoroughly discussed, I agree with most of what most of those authors have said (excepting Newberg, because I didn’t find anything reasonably accessible that gave a good view of his thoughts). I never once argued that religion does not have benefits. But, again, the best available evidence does not find that religion has greater benefits than a good social club, and combined with the obvious problems with religion (maybe not *your* religion – but the word is not limited to what *you* believe), religion does more harm to society than good. Nothing you’ve presented (and nothing I found whilst checking my own position in order to engage with you as fairly as possible) seems to contradict that position – and I don’t pretend to be an expert, but when I find I’m largely agreeing with the people you’re citing, I feel I must be somewhere along the right track. If my understanding of the situation is consistent with the evidence offered by the experts *you* cite, I feel rather comfortable with my claims.

    3) I’m rather sure that you offered up Bayesian analysis early in the conversation as the basis for forming beliefs (something I agree with). “God” is the most extraordinary claim imaginable – under Bayes’ theorem, it *should* be the hardest claim to prove. Beyond that, though, I think you’re still misinterpreting my stance. It’s not that a different type of evidence is required – just a heck of a lot more of it, and certainly more than what’s available right now.

    But to argue, in effect, that I need to believe in order to see the truth is an absurd standard that I doubt you apply to *any* other claim you deal with – and is yet again hypocritical in light of your accusations of my epistemology being circular. Do you need to believe in ice cream before you can discover that ice cream exists? Do you need to believe in love before you can feel love for another person? Evidence should stand alone and lead to a conclusion – we shouldn’t need to assume the conclusion is true in order to establish the truth of the conclusion. If a reasonable person cannot take a fact (or collection of facts) about the universe and get to god *without* first presuming that god exists, then the evidence in question isn’t worth talking about.

    As you’ve noted, we’ve obviously reached a point of diminishing returns in the conversation. I would sincerely recommend that you examine your epistemological standards, and attempt to genuinely understand why atheists think that your standards are flawed (hint: it’s very likely the same reason that you, presumably, think Muslims, Hindus, Buddhists, Jains… have epistemologies which lead to false conclusions). I would once more state that, in your original post, you’re unjustifiably attempting to shift the burden of proof onto atheists (who by definition lack belief, and are not making claims about what does or does not exist – and therefore have nothing to defend with respect to the question). Otherwise, I’ll thank you for what was, up until this final interaction, an interesting and largely positive discussion, and wish you well.

  54. Hi Cheomit,

    As I said, I have read your response, and I am sorry you think I’ve been a hypocrite when I think I have been quite consistent. I still think you are responding to what you think I am saying and not to what I am actually saying – there were several points when I quite strongly felt like correcting your misunderstandings but I think we are past the time for that.

    So as I foreshadowed, I will draw a line under this discussion now, certainly before either of us become more frustrated with the other. Thanks again for your courtesy and for sharing you views.

    Best wishes to you, Eric

Comments are closed.