I've just returned from a wonderful 2-day philosophy of religion workshop at Glasgow organized by Victoria Harrison, who put together a diverse and high-quality program. One of these exciting papers was by Joshua Rollins' (U of Oklahoma) on the common consent argument.
Roughly speaking, in its crudest form, the Common Consent argument (CCA) goes as follows:
- Most people believe in God
- Therefore, God exists
The proposition that the mere popularity of a belief might constitute evidence for its truth may strike us as odd. Mill, for instance, argued that common opinion might be OK for the common folk who are unable or don't feel entitled to form their own opinion, but to us, thinkers "the argument from other people's opinions has little weight. It is but second-hand evidence; and merely admonishes us to look out for and weigh the reasons on which this conviction of mankind or of wise men was founded".
Kelly regards this Millian view as implausible, especially in the light of recent social epistemology (see above).As for common consent, he observes is often reasonable to defer to majority opinion. An example he offers (and I encountered in real life many times) is the following: suppose I think I need to put out my bin on Friday as always, but I notice everyone else's bin standing outside on Thursday. I will, justifiably, revise my beliefs and put my bin out too (I usually only afterwards realize there's a bank holiday upcoming or some such, but by then I have already deferred to the majority view). Social epistemologists recognize this importance of other people's beliefs as prima facie evidence for the truth of these beliefs. Thus, given that a large percentage of the population are theists (recent estimates somewhere between 70 and 90%), we come to the datum that Kelly puts forward as the basis for the CCA:
Datum: A strong supermajority [i.e., more than 60%] of the world's population believes that God exists.
On the basis of the datum, one can construct an updated CCA to the effect that the datum provides (defeasible) evidence in favor of the proposition that God exists. There are several counterarguments Kelly discusses. I will here sum them up briefly, but will not go into the replies he offers on behalf of the theist:
- bite the bullet (i.e., say that the datum provides evidence for God's existence, but that on the whole, evidence such as gratuitous evil points to atheism)
- argue that the unsophisticated are overtly represented among theists, whereas the well educated/critical reasoners etc. are more likely to be atheis
- argue that the datum is insignificant, because people did not come to their beliefs independently
- argue that the datum is false - religious beliefs do not have the same referent (i.e., Allah and God do not refer to the same being).
Kelly thinks (3) is the most significant problem, and I'll look here at what cognitive science can say about it. Imagine the following possible worlds. Each of them is similar in that a supermajority of the population believes in God. However, how they come to these beliefs differs:
- w1 - people come to belief in God independently, through critical reflection (e.g., consideration of natural theological arguments)
- w2 - people come to belief in God through cultural transmission (e.g., from their parents)
- w3 - people come to belief in God because it's innate
- w4 - people come to belief in God as a combined result of cultural transmissions and some innate cognitive predispositions.
Intriguingly, Kelly thinks people who live in w1 have the strongest evidence for the proposition that God exists. He thinks w3 is a lot less strong, because it is vulnerable to defeaters (e.g., cognitive explanations of why we believe in God such as adaptationist accounts that say such a belief is evolutionary advantageous). This is interesting because many CCAs (e.g., Calvin's) have the claim that belief in God is innate. Such a claim isn't necessary for CCA and, according to Kelly, doesn't help the theist.
I think we probably live in w4. Religious belief results from a combination of what cognitive scientists call content biases and context biases. Context biases have to do with the context of testimony, e.g., peer pressure or parental upbringing. Content biases have to do with the content of the testimony - what is the actual thing being testified. Cognitive scientists have convincingly argued that religious beliefs are easily acquired. For example, Pascal Boyer has argued in his Religion Explained that there is something attractive about counterintuitive ideas like listening statues and invisible people. We tend to remember such ideas better than boring ideas like inanimate status or visible people. Hence the attraction of religious beliefs.
It's clear at this point that not all such cognition and culture explanations will debunk CCAs for a variety of other beliefs. After all, a combination of context and content biases can account for the widespread human belief in the existence of trees, cats and other minds. The fact that our minds, collectively, hit upon such concepts, provides in such cases at least prima facie evidence for them. But when we delve in a bit more detail into the evolutionary explanations for such beliefs, we can see why they aren't defeaters (i.e., widespread belief in trees is best explained by the actual existence of trees, on which our evolved capacities latch on, whereas widespread belief in God is commonly not naturalistically explained by invoking an actual deity - I'm not saying such a supernaturalistic explanation is not possible, just that it's not on the table in cognitive science of religion).
Cognitive science thus seems like a viable way to undercut CCA. A lot of the standard moves philosophers make when responding to debunking arguments, and that work reasonably well for justifying one's individual religious belief in the face of such defeaters, don't work so well in the context of CCA. For instance, some theists respond to evolutionary debunking arguments by saying a naturalistic & supernaturalistic explanation aren't competing, since God works through natural selection to bring in us about a belief in him (a sort of thomistic account). Although a CCA proponent might take this line, it does, I think, take the bite out of CCAs.
Perhaps the best line of response for the CCA-defender is to say that such explanations aren't on the whole that impressive. Cognitive science of religion is still in full development, they haven't even reached consensus on whether or not religion is an adaptation or byproduct, many features of religion remain unexplained (e.g., mystical experience, for instance, is studied but not really naturalistically explained in CSR). Moreover, relying on the pessimistic meta-induction, combined with the immaturity of the scientific study of religion, a theist can argue that today we aren't particularly impressed by past debunking arguments such as Freud's, Marx's or Feuerbach's (although they continue to be reiterated in theology and philosophy, no cognitive scientist takes them seriously, except perhaps Freud's with a lot of modification, see e.g., Kirkpatrick). I'm not sure if these movesÒ will help the theist. In any case, the fact that CCAs seem - at least to me - somewhat vulnerable to CCAs indicate that best explanation considerations loom at the background of CCAs.
For instance, the best explanation why people put out their bin on Thursday in my earlier example is that the bins will in fact be collected on Thursdays (there are other explanations such as that everyone in the street but me is victim to some practical joke, but these seem less plausible). But a naturalist will not accept that the best explanation for the datum is that God exists. She could cite cognitive science explanations as naturalistic explanations for the datum (This post is a bit longer than I intended. Anyway, I'm now returning on a long, but eco-friendly train journey from Glasgow to Oxford, so I won't be able to respond to comments for a while, but it would be interesting to hear if others' intuitions might differ from mine.)
Wouldn't this percentage be much higher (and therefore more compelling) if we consider the (probable) percentage of theists over the course of humanity's existence? The other advantage of this is that it removes the temporcentric egotism so prevalent in discussions of progress of knowledge.
Removing the contemporaneity from the argument as reduces the power of the argument
when you consider individuals such as (warning: grossly European and Catholic in bias) Augustine, Anselm, Abelard, Bernard of Clairvaux, Maimonides, Averroes, Thomas Aquinas, Martin Luther, Samuel Johnson, etc, and almost all educated individuals until the "Enlightenment."
On a personal note: the CCA argument, in a much cruder form, was one point of conversion for me. As above, I had figured that over the course of human existence, the greater majority of people had believed, and as most people mostly think with common-sense reasoning were likely correct about the existence of a transcendent spiritual realm/existence. Secondarily, (and not necessarily logically following) I came to the conclusion that atheism had the tinge of an elitist and paternalistic point of view. This realization to my mind was pretty compelling and added to the growing list of reasons why I should be searching out a practice of faith.
David, thanks for these observations. I agree that (2) theists are (presumed to be) less sophisticated is not a big objection for CCA, except perhaps for people who want to conjointly hold a progressivist view. And even then, there is still a significant percentage of theists. If we restrict it to specialists, who have done a lot of careful thinking in this field, I find the fact that between 72% and 74% of philosophers of religion are theists prima facie defeasible evidence for theism. By this, I do not mean it's incontournable, but it's something the atheist/agnostic will have to explain (away).
Helen (if I may): Interesting post - Thanks
Coule of observations:
1) It seems to me that the social epistemologist has to have an error theory to explain those who disagree with the super majority. What is it? If the explantion is some form of coherence theory like Rawlsian considered judgments, then it seems that are in a position that the p and -p can both be true depending on the perspective of the person making the claim. I do not see how this helps determine if there is a God if we are trying to make some ontological claim.
2) Regarding your bin example - it might be that you lack some information that the others have. For example, you forgot that Friday is a holiday, or you did not read the notice that the pick up dates had changed. Had you possessed this information then you would have put your bin out. It seems to me that if one were in the position you describe that one would put one's bin out, but inquire of others as to why they put their bins out. CCA might motivate us, but it seems to lack ongoing justificatory force to continue to believe in putting bins out on Thursday. It would seem to be beter epistemic practice for us to seek confirmation other then CCA in this type of situation?
Hi John: thanks for these comments. I'm not sure if a coherence theory approach would allow for disagreement of this sort (e.g., I've just reached a different considered judgment, based on my stock of belief). For again we can ask: why does the minority reach the different conclusion?
Regarding the bin example: I framed it in such terms that - lacking all other information - I seem entitled to put out my bin. Perhaps I'm living in a sleep town and everyone's off to work already, and I can't ask them, I just have the visual evidence that the bins are standing outside. I don't have any clear intuitions about what I would do Wednesday night--I think I would do what you suggest and try to find out beforehand if the dates changed permanently. Here's the intriguing thing: we rarely ask others why they are religious believers/agnostics/non-believers. Perhaps the question is trivial, since it's already settled: people are religious believers because their parents were, and they were brought up into it (problem 3 in the post). However, even here Kelly insists widespread theistic belief - by itself - has evidential force because those people continue to believe.
It seems obvious that if the majority of people believes p, that's evidence for p.
Here's a Bayesian argument. Let p be a proposition about which we have no information, and let q be the proposition that the majority of people believe p.
Then, either p and q are probabilistically independent or not.
It is absurd to think that they are independent. First of all, in the real world, we almost never, and perhaps never, meet independent propositions. A standard example is something like subsequent results of a coin toss. But those aren't really independent--each time a coin falls, the metal gets worn asymmetrically, and that affects the result of the next throw. Second, it just seems implausible to suppose that there is complete independence between truth and belief.
So, if p and q are not independent, either they are negatively correlated or they are positively correlated.
If they are negatively correlated, then the fact that the majority of people believes something is evidence against the proposition. I suppose some misanthropes believe this, but it seems quite implausible.
So the remaining option is that they are positively correlated, and hence the fact that the majority of people believes something is evidence for the proposition.
Helen - Thanks for your reply.
1. 'For again we can ask: why does the minority reach the different conclusion?' Hence the need for an error theory. If such a theory were provided would that not supercede CCA?
2. In order for us to accept CCA as evidence of p do we not have to have some basis other then CCA to warrant holding that if the majority of people beleive p then that is evidence for p? For example, in the past we have found evidence that supports belief claims that are not dependent on the majority believing p? This applies to your bin example in so far as you would have some other basis for thinking the majority correct - e.g. they have been proven correct in the past. Of course, maybe CCA is simply a way of formulateing a 'herd' mentality.
3. If CCA is reasonable then it can be used against theism, or at least some aspect of theism, in so far as the majorty of people do not think that, for example, Catholicism is true. It just seems that CCA is a non-starter in that it seems a very weak epistemic position to maintain that p is true because the majority think that it is true. This epistemic weakness does not lessen if there is lack of any other evidence for p. If CCA were sufficient to warrant belief in absence of other information then those that do not think that p is true - those that take the minority position - would have to admit they are wrong and change their beliefs. But that does not happen. The problem seems to be that if we accept p being true in absence of evidence other the the majority thinking that it is true then what would we take to be evidence againsn p? Are we falling back into 'Meno's paradox?'
John:
Let me try another thought experiment. You wake up, with no memories of anything, walking in the middle of a long line of people, on a narrow path suspended over an abyss. The path forks ahead, and the two paths going out from the fork, still over the abyss, are shrouded in mist. There is a loud announcement: "Remember that one of the paths leads to good place and the other to a bad place." You notice that the majority of people go to the left at the fork. Neither the people who go left nor the people who go right look like they're guessing--they look they believe themselves to be going to the better place. You ask the person behind you and in front of you what's going on, but they don't speak your language.
What do you do? You can't stop because of the press of people behind you. You must choose--left or right? The only data you have that differentiates left from right is that most people go left.
And surely that's exactly what you should do.
Mr Pruss,
When you add the promise of reward to the thought experiment, I think you're no longer so much in the realm of the CCA but rather Pascal's Wager. Your example (while a vivid story) is more about hedging your bets for the afterlife than knowledge and faith in the here-and-now.
I don't know if that thought experiment works out so well. Under the hypothetical's assumptions, why suppose P(Left is good|Majority chooses left) > P(Left is bad|Majority chooses left), absent some assumption regarding the typical behavior of humans (which, lacking memories of anything, we would not have reason to think)?
John: CCA as formulated today is not a straightforward inference from a majority opinion to theism. It says that the majority opinion provides defeasible evidence for theism. An atheist could acknowledge this, but nevertheless stick to her guns and say that all evidence considered, she thinks atheism is more likely *all things considered*, e.g., by relying on evidential arguments from evil. I don't think it would be good epistemic practice to simply always go with the majority opinion, no matter what (that's herd mentality, as you point out).
However, this does not take away the fact that the majority opinion does constitute evidence of some kind.
Hi Helen,
I was hoping you might clarify a few things. When I see the CCA, my first reaction is to think that the argument would be good if the premise showed that the conclusion was likely to be true. If (1) isn't any sort of evidence for (C), it's hard to see how the CCA could be a good argument. Even if (1) is evidence for somebody for (C), however, it still seems that the argument might be no good at all. We might say that (1) constitutes evidence for S for (C) if (1) raises the evidential probability of (C) to some degree. It's hard to rule out the following possibility: there's somebody with a body of evidence such that adding (1) to their evidence means that (C)'s probability gets some boost. Of course, it's also hard to get excited about _that_ possibility as that possibility might be one in which the boost is vanishingly small. So, learning that my mom mailed me a lottery ticket, I acquire evidence that I'll be rich, but nobody would think there's an argument worth consideration that takes as a premise that my mom mailed me a ticket and offers that premise in support of the conclusion that I'll be rich.
Also, I'm genuinely curious what people would say about the parallel argument?
1'. Most people believe Jesus isn't God
2'. Therefore, Jesus isn't God.
The two seem parallel to me, but I can't imagine anyone being seriously troubled by this argument. I can think of a number of plausible responses that Christians might offer to try to undercut the argument, but they seem equally available to the atheists who want to resist the CCA.
This second argument does raise an interpretive issue. I can imagine a Christian saying that we'll see that (1') is false in the fullness of time. Should we think of all people at all times or some more restrictive set? I'd like to think, as an atheist, that there's a brave new world coming in which everyone accepts (1') because we've all embraced atheism and we do the necessary deductions. Then, you know, the world becomes uninhabitable, we spread across space and take our godlessness with us while we colonize the distant planets. And then it comes to be that most people believe that there's no God (our future planet hopping atheists outnumber the theists that died off with Earth). If that happens, have we made the first premise of the CCA false or in evaluating what 'most people' believe, should we focus on people up to now? To my mind, that exclusion seems arbitrary. (Thanks to Sam Harris and Richard Dawkins we're all better at evaluating religious claims. (That's not intended to be a factual statement!)) If we drop the restriction to include all people past, present, and future, it's hard to see how we can be in any position to determine whether the argument's one premise is true.
Alex's thought experiment sounds a bit more like William James than like Pascal to me. I suppose that majority opinions and lifestyle choices to play a role when we make pragmatic decisions as well. Still, I think Alex did not mean this as a metaphor on afterlife beliefs, more something about practical considerations when making choices about world views.
Thanks, Clayton. I'm not sure whether it's in this case productive to phrase evidence in terms of probability, which is what you do in "We might say that (1) constitutes evidence for S for (C) if (1) raises the evidential probability of (C) to some degree." (I know many people do it this way, but it is not the only way to look at evidence).
The strength of the evidence is a different matter. A successful CCA would have to say that the evidence of general consent is not small or negligible, all things considered. I assume at best it could be part of a cumulative case for God's existence (e.g., taking into account other evidence that might favor a supernaturalistic or specifically theistic worldview).
The parallel argument and worries of atheists taking over the globe are interesting. They call to mind William's concept of knowledge, which invokes a form of reliabilism that relies on safety-from-error. If the public opinion is apt to change for all sorts of reasons (e.g., ideological success of Dawkins, more due to his writing style and PR than actually the cogency of his arguments), we aren't safe from error when we take the majority opinion since it can change at any given time.
So perhaps the person who accepts the majority opinion should take into account facts about the perceived reliability of these agents. I think there is room for disagreement here, similar to reductionist and anti-reductionist positions in the epistemology of testimony. Applied to this case: I could say that I am prima facie entitled (which is weaker than justified, but still means I'm within my epistemic rights and not being irrational) if I take majority opinion that p as prima facie evidence for p, even without considering how these people formed this position (an anti-reductionist stance), or I could insist that even though majority opinions can be sources of evidence, I am only entitled to treat them as such if I have some additional positive reason for trusting them, e.g., evidence that the opinion was formed independently by each member (a reductionist view). It seems to me CCA relies on some form of anti-reductionism.
Everybody:
CliveStaples:
I agree--one does need to suppose that one remembers certain general features of the world, like that these two legged bipeds are people, that people tend to seek to benefit themselves, etc.
Mr Mayeux:
I think you get more than a pragmatic conclusion from the case. One concludes that it's better to go where the majority goes, absent further information.
If rationality requires that we go left, then we had better have assigned a higher credence to the claim that the left path leads to the better destination. After all, the utility of going left is gP(left is good) + b(1-P(left is good)), where g is the utility of the good destination and b is the utility of the bad destination, and the utility of going right is bP(left is good) + g(1-P(left is good)). Assuming g>b, it is easy to see that it's worth going left if and only if P(left is good)>1/2.
So if rationality requires that we go left, as it does, it must be that given the evidence, it's more likely that left is good.
But what is that evidence? Surely the only evidence we have here is that most of the people think that left is good.
So that most of the people think that left is good is evidence that left is good.
Alex: I do not disagree with Helen that we would put out our bins based on CCA. I am simply tring to determine the epistemic value of CCA. To your thought experiment: if I decide to do what the majority does, what do I base this on? I must have some knowledge (memories) of some adage refercing CCA. I do not see how I can fulfill the criterion of having no memories.
Helen/Alex: Why would we not simply limit our response to S (either individually or as a group) believing p to be true to simply asserting that 'S believes p is true' without assigning any epistemic status to p itself? If we were ever in a situation without any other information then CCA would we not seek out additional confirmation regarding the epistemic status of p? It seems that that would be the epistemically prudent position to take. I would put put my bin, but I would not do so simply because others had done so, but because I thought that I lacked some information that they possessed and I would seek out that information to confirm my decision. It is the belief (information) that I lack information that actually is the deciding factor in my putting out the bin. To go out on a rather far, and possibly thin, limb, it seems to me that CCA is a disguised divine command theory.
Question - if we agree that we would put our bins out on Thursday instead of Friday, as we normally do, if we saw that others had done so, would we put our bin out 1st on the next Thursday? It seems to me tht we for predicting what we would do. We do use some other criteion, even in the orginal bin scenario.
At the beginning of the 20th century virtually all educated people, theist and naturalist alike, believed that physical reality is deterministic. What’s more they were right in believing in physical determinism on the evidence available to them. But they were wrong, for today the evidence is equally overwhelmingly against determinism.
Today astrology is probably believed by more people than theism – and more strongly.
My point is that the relationship between common consent and truth is so volatile that the whole question is rendered inscrutable. I think the relevance of the CCA lies only within the context of the idea that belief in theism is basic.
Incidentally, the fact that the human condition is religiously ambiguous can, on theism, be explained thus: God wants morality to be prior to rationality, and thus built a world in which foundational belief follows moral commitment, and not vice-versa. In other words, God values the free choice for the good especially at the absence of knowledge. Indeed, where is the merit in choosing to follow Christ when one knows that Christ is the truth?
I'll side step the cognitive science responses for a moment just to mention that there are lessons in the epistemology of testimony that bear on the independence issue. A good place to start reading is Lackey's "Disagreement and Belief Dependence". Lackey argues pretty convincingly that even in an case were everyone shares the same first order evidence there can still be independence.
Helen, interesting post.
My two cents:
When it comes to assessing whether there is common consent for a belief, shouldn't we count whether the meaning is the same, rather than the referent?
I'm asking because of one of the objections discussed by Kelly:
Obviously, real beliefs are far more complex, but to explain what I'm getting at, we could have a scenario in which:
a) 21% of the world's population believe that there is a designer of humanity (say, belief B1), but do not believe either B2 or ¬B2, and do not believe either B3 or ¬B3.
b) 20% believe that there is a necessary being (say, belief B2), but don't believe either B1 or ¬B1, and do not believe either B3 or ¬B3.
c) 20% believe that there is a being that is the most powerful being of all actual beings (say, belief B3), but don't believe either B2 or ¬B2, and do not believe either B1 or ¬B1.
Then, if there happens to be an actual being that matches all three criteria, their beliefs have the same referent, and so there is over 60% of the population who have beliefs with the same referent, which is a necessary being, designer of humanity, who is the most powerful of all actual beings.
However, there is no common consent for B1, B2 or B3, let alone for the conjunction of them.
So, in order to assess whether there is common consent for a belief, it seems to me we should consider meaning, rather than referent.
Another issue I'm considering is the following:
Okay, that sounds like a good explanation without any further data.But what would happen if you have 1/7 of the population who put out their bin on Monday, 6/7 who tell them that they're wrong because nothing happens on Monday, then 1/7 who put out their bin on Tuesday, 6/7 who tell them that they're wrong because nothing happens on Tuesday, etc.?
In that case, it seems to me that without further information about this peculiar behavior, we wouldn't have any good explanation for what's happening, and it's also not at all clear to me that the least bad explanation for such behavior is that some day of the week the bins will in fact be collected, but people for some odd reason disagree (often strongly) about the day; rather, it seems to me that we wouldn't have enough information to assess that matter without further data, either (like taking a look at whether bins are actually collected).
Let's suppose that you also have an overwhelming amount of evidence that shows that, historically, most humans hold beliefs about bin disposal that are false, and moreover clearly false in the eyes of nearly all of the population who come from very different societies.
There was an Indian guru, Sathya Sai Baba, who claimed to have the power to manifest small objects (i.e., he could just make them appear), and his followers bought into that. It seems that his followers were over a million (from en.wikipedia.org/wiki/Sathya_Sai_Baba).
There was also Sai Baba of Shirdi (from en.wikipedia.org/wiki/Sai_Baba_of_Shirdi).
Of course, we know that such claims aren't true.
We may also consider a Catholic example (for instance; en.wikipedia.org/wiki/Januarius ).
I do not know how much agreement I could get here, but the substance does not become liquid, or it's not merely blood (we know blood does not do that), but plenty of people believe it (though most people would actually reject it, as long as it's incompatible with their religion).
Those are merely a couple of examples, but my point here is that when it comes to superhuman powers, there is overwhelming evidence that most humans historically (and even today) hold false beliefs (we can often tell that just by mutual incompatibility of different beliefs), and that such beliefs are often held passionately and firmly.
The same goes for beliefs about design and/or designers: Regardless of the issue of compatibility between evolutionary theory and certain religions, the fact is that in the past (and maybe today), most people had different beliefs about specific acts of creation, which are incompatible with a process of evolution that has lasted for billions of years so far.
P.S: This captcha is really difficult. I almost gave up trying, and I'm not sure I'll be able to post further comments.
AM : thanks for these comments, and I apologize for the captcha problems.
I am wondering if CCA could only be defended weakly (something like: most people in the past and today have believed in some form of supernaturalism, this fact is evidence for supernaturalism being true)? Or could we get at a more specific version that would be more congenial to theism?
Take the creation example: across history, people have held different beliefs about creation (ex nihilo, from pre-existing matter, by one person, a committee). At most a CCA could say that's evidence prima facie for the view that (some aspects of) the world are created. A theist would probably need additional considerations, for instance using simplicity, to say it's one person rather than a committee, ex nihilo rather than from pre-existing matter.
The miraculous stuff you point at seems to not apply to CCA, since the relevant class of people are not just those who believe in the miracles but people in general. Otherwise, the CCA would be pretty much circular. As you point out yourself "most people would actually reject it [the miracle], as long as it's incompatible with their religion".
Helen, thanks for the reply, and no need to apologize; btw, I'm thinking it may be a problem on my end (old monitor).
Also, I see what you're trying to say, and sorry if I've been unclear in my reply. I'll try to clarify my view on the matter:
Personally, I prefer not to use terms like 'supernatural', and 'miracle' (and by the way, 'naturalism', etc.) for a number of reasons, but leaving that aside at least for now for the sake of brevity, and while I do not know how you precisely distinguish between actions by entities of superhuman powers that are miracles and those that are not miracles, chiefly the point I was trying to make in that part of my post is that humans tend to usually make false claims in categories that seem to be relevant to assessing the ACC, and that that essentially at least cancels any prima facie support that common consent among most humans in a random category of beliefs would give to a certain belief.
More precisely, I'm considering categories of claims like the following.
a) Claims of existence of entities of superhuman power.
b) Claims of actions by entities of superhuman power.
c) Claims of acts of design by entities of superhuman power.
We know that at least nearly all claims in each category are false. Those are not generic claims of a designer, but specific claims about, say, Pan-Gu, Zeus, Quetzalcoatl, etc. Millions firmly believed (or believe, depending on the entity) in the existence of such specific beings, and that they had carried out such actions, etc. But they were (and are) all wrong.
Given that claims in such categories are usually false, even if you could extract some common points between them, that would not really provide any non-negligible support for the part that is common, but rather would call for studying the causes of such human behavior before we can go any further.
But now I see that perhaps I'm being unclear about what I'm trying to say, so maybe the following analogy would be more clear:
Let's suppose that the police find a human body in the woods. It's only a skeleton. But they do not know anything else at that point. So, they ask any witnesses to help, and the following witnesses show up, providing the following accounts:
Alice says:
Yes, I know what happened. I saw everything:
Six men got into a brutal fight with twelve other men, and one of the six was beaten to death. The other five managed to run away as their friend was killed.
Bob says:
Yes, I know what happened. I saw everything:
A woman was just walking around, minding her own business. I saw another woman approach her, and shoot her at point blank. It was a single shot to the head, and then she left as if nothing happened.
Tom says:
Yes, I know what happened. I saw everything:
A man and his wife were running from three gang members, screaming for help. One of the thugs caught up with her, and her husband tried to defend her. But the thugs stabbed him dozens of times. Then, they kidnap her. I do not know what happened to her after that, but I know they got her.
Mary says:
Yes, I know what happened. I saw everything:
Two men got into a heated argument, until one of them let his three Rottweilers loose on the other. They just tore him apart.
Thousands more show up, and the accounts remain just as vastly different.
Given that information and nothing else, the police should not conclude that the body resulted from a homicide, or even that they have good reasons to suspect homicide (rather than, say, accident) that they did not have just by finding the body and in absence of such notoriously unreliable witnesses.
Still, at least normally, when many human witnesses claim that there was a homicide, there usually was one.
It's an analogy and not a perfect match, of course: for instance, false claims in the categories a), b) and c) above are not unusual at all but common behavior. However, that difference only works against the ACC as far as I can tell, since it's not even the case that normally, claims in such categories are true. Normally, they are false, even if deeply held and seriously defended.
So, in the previous scenario, it could be argued (though it's still quite debatable) that at least the police would have some reason to suspect that one of the witnesses is correct, even though there is surely not conclusive evidence or anything like that, since that piece of evidence is seriously undermined due to the odd case of vast unreliability surrounding that specific case.
But in the case of claims in categories a), b) and c), they're normally false, so that does not seem to work, either.
Take the creation example for instance. Some people held or hold that Pan-Gu created some aspects of the world, but he (the first creator) came from an egg (i.e., mindless stuff is original, not mind). Others believed that Coatlique was impregnated with a certain knife, giving birth to a number of other beings, including the stars, etc.
However, we know that those stories were all false, that most religions have become extinct, and that the claims of existence of such designers of superhuman powers, were false.
In my view, the fact most humans are prone to deeply held beliefs that are not even close to reality (namely, beliefs in those categories) calls for an explanation based on human psychology (including sociology, etc.). Given that errors are so common in some categories, one could expect some underlying general features of human psychology that makes humans prone to incurring errors in those categories, and that would also explain the similarities (i.e., it's in those particular categories in which (normal) humans are prone to some kinds of errors). But in any event, if you have a case in which the normal claims in a category are false, that undermines the common consent idea.
Of course, someone might argue that those aren't relevant categories, and then try to come up with a modified ACC, but it seems to me they would need to get clear about what specifically there is agreement about, why the previous categories (and similar ones) are not relevant, etc. As it stands, the ACC fails as far as I can tell.
Very interesting take on CCA Helen, so thank you for that. My own disclosure I actually do believe in a God, what that means, I could not say.
The problem I have with CCA can be illustrated by periods in history where racism of some sort or cultural blame was the dominant belief system throughout the globe. Pick a time like WWII, slave times for African Americans, Roman enslavement and pillaging, etc...
Psychologists often describe "mob mentalities" or group thinks that extend to a wider social adoption of a given belief of set of beliefs that later prove to be false or inspired purely by ignorant fear.
Does fear of death/the unknown inspire a belief in God, or ignorance of why/how we are here lead to filling of the blanks where a God is invented?
Anonymous:
The fact that majority opinion does often go awry (as in the cases you describe) constitute a worry for CCAs, since it seems that one's beliefs, even if true, are just true as a matter of luck (e.g., suppose I were to abhor slavery for no other reason than that most people today abhor it, my belief if true (supposing moral realism, etc etc) would just be lucky because I would have believed differently 2000 years ago.
So to refine an updated CCA a bit
1. If the majority of people believe that p, that's prima facie evidence for p
2. A supermajority believes that God exists
3. Therefore, we have at least one prima facie piece of evidence for the existence of God.
Now, we have discussed possible attacks on premise (2) (see also comments above), but I think there might be a way out, e.g., say that most believers have been and are theists (whether monotheists or polytheists), or weaken CCA in favor of supernaturalism.
I would add that (1) is only plausible if the belief has some stability across time. Otherwise, the luck element comes in. So for instance Kelly briefly discussed the fact that people across time and cultures are realists about the existence of the world is prima facie evidence for this claim. I think this can be granted for theism, or at least for supernaturalism.
Second, against (1) we can level undercutting defeaters. I've read a lot of the literature on terror management theory and anthropology, and am inclined to believe that fear of death is not a motivating factor for theism. Although there are small effects, such as an increased acceptance of creationism when primed with reminders of one's own mortality, overall the connection is not overwhelming. Younger people are more afraid of death than older people, but paradoxically, older people are more religious than younger ones (this, by the way, is a cohort effect - it's not the case that older people become more religious as they grow older, as a longitudinal study by Bengstrom has shown). Also, there are religions with an unattractive afterlife or no afterlife where belief in gods, ghosts etc is prevalent (if you read Ecclesiastes, for instance, you can see how belief in God is independent from belief in an afterlife).
The current cognitive science however, and I think this is a problem for theism, doesn't invoke God as an element in the explanations of why we come to hold religious beliefs. By contrast, for realism, a plausible explanation of why belief in it is widespread is that realism is in fact true - animals that were idealists or solipsists simply did not do so well in surviving and reproducing.
AM: Hmm, let's see. I see what you are getting at.
The idea you propose is roughly this:
1. people across time/cultures have commonly believed in miracles/creationism etc.
2. almost all of the *particular* miracle stories, creationist stories etc. have turned out to be false
3. therefore, probably there is a good undercutting defeater (sociological, psychological explanation for these beliefs
4. therefore, CCA fails.
Going from 2 to 3, you need a lot of extra assumptions (e.g., "Given that errors are so common in some categories, one could expect some underlying general features of human psychology that makes humans prone to incurring errors in those categories, and that would also explain the similarities (i.e., it's in those particular categories in which (normal) humans are prone to some kinds of errors").
Take a different example on witnesses: there are a number of witnesses, all of whom agree they have seen a car crash on a given location. Here, there is not even material evidence that a car crash took place. The witnesses don't agree on anything else but that they saw a number of vehicles collide on a given road. Some say it was a van and a car, some others say they were two station wagons, there is no agreement on the color of the vehicles. A minority say it was a bicycle and a car. My intuition here is that the police should take these reports, even though they wildly differ, as prima facie evidence that a crash involving a number of vehicles took place on that road, even if there is no further material evidence for such a crash at all. While it may not be enough to warrant further police investigation, it still looks like prima facie evidence.
Helen, thanks for the reply:
It was more like the following (though I was considering more than one category, but to keep it short):
1. At least the vast majority of claims about the existence of beings of superhuman power made by humans across time are false.
2. At least the vast majority of humans who make claims of existence of beings of superhuman power, make false claims.
3. Given that, it seems human majorities are not a reliable guide to the existence of such beings, or provide non-negligible support for that.
Something similar can be said about designers, etc.
Yes, but I would say that that is because when many humans make a claim that there was a car crash, there usually was one. On the other hand, we also know that when humans make a claim that an entity with superhuman powers exist, they (at least) almost always are wrong, even though they are highly committed to such beliefs.We don't need the extra assumptions, by the way. I added some extra points because they're plausible, but the above should suffice as a serious undercutting defeater.
I'd add that they did not do well precisely because realism is true.
On the other hand, it seems to me that, plausibly, early humans who rejected the common consent in their society about the existence of certain beings who engaged in specific actions, the performance of certain rituals to appease or otherwise communicate with those beings, etc., plausibly also did not do well in surviving and reproducing, even though all those religions got extinct, and all those beings are non-existent. Of course, the reason for their lack of reproductive success in that case has to do with the reaction of other people, not with the actions of the beings of superhuman powers in question, who turned out not to exist.
Helen thank you for your response. I for one, reject terror management theory. I have peers who have performed research using this theory as the basic framework in their research and have also studied the theory myself. I do not think it applies to the real world.
Regarding theism, Cognitive Science in my opinion cannot never do any damage to theism, but it can demolish the CCA theory.
As far as older people being more religious and less afraid of death, how was this measured, by self report surveys? I am not saying it cannot be true but I would have to see well constructed peer reviewed research with sufficient sample sizes and randomized methods to comment on the claim. Even if it is so, it still seems religion can play a role in alleviating fears of death since fear of death is,a universal human feeling though some have more overt fears than others.
I think CCA is on the chopping block but not theism. I of course respect your opinions and am writing a response since this blog and your writing on this matter do interest me.
Helen,
I've been thinking about your 'potential accident' scenario with more time, and I realize I may have underestimated the degree of agreement between the witnesses; so, now I think that maybe we don't need to know that when many people say there was a traffic accident, there usually was one, but is enough that 'traffic accident' is not a suspect category (i.e., we do not have any good reason to believe that when people make claims about traffic accidents, they usually did not happen).
So, when someone says there was a traffic accident, the police have prima facie evidence. As more witnesses show up, they also say that there was an event at a given location (and, presumably, they agree at least on the date, so at a given time too), so the discrepancies may simply be the consequence of the fact that people usually do not remember details so well.
On the other hand, if the discrepancies become too big, that surely undermines the evidence, so it's a matter of how big those differences are (i.e., is it still the best explanation that they're describing an actual event?).
And if the police have found no forensic evidence but they didn't look for it, that does not affect the evidence. On the other hand, if they look for it and do not find it, that's surely a problem. And if they keep looking with no luck, then they should conclude that there was no accident.
As I see it, beliefs used to try to back the ACC are different in a number of relevant ways, such as:
a) There appears to be no good reason to think, even preliminarily.
For instance, suppose Chen says that in the beginning there was only chaos, and then an egg was formed, existed for 18000 years, then Pan-Gu was born, separated Yin from Yang, created the Earth and the sky, etc., and eventually died and different parts of his body became different stuff (rivers, etc.).
Later, Nüwa created humans.
But let's say Bob tells you that a being that is three persons created the universe in seven days, in such-and -such ways, etc., created humans, and the like.
It seems to me that there is no good reason to think that they're merely describing a real event differently; rather, they seem to be talking about very different entities, events, etc.
b) Even if there is some similarity, the categories in question (see my posts above) seems to be suspect, since we can tell (see above) that nearly all claims of actions (or even existence) of creator beings of superhuman power, etc., are false.
Also, that's not a made-up category: the reports that would allegedly support the ACC are precisely the claims about Zeus, Thor, Pan-Gu, Quetzalcoatl, etc.
Anonymous: thank you for the response.
The study, which is as far as I know not published, was longitudinal (Bengstrom is the lead researcher-. They started out with a sample of about 2000 people born in various times in the 1970s and asked them a series of questions (e.g., about their belief in God, prayer frequency etc). Every decade or so, they asked the same people the same questions. They found that the main predictor of religiosity was age cohort (with people born earlier more religious). Some fluctuations occurred over time, i.e., some religious people became non-religious later in life and vice versa. But the net effect was not statistically significant. My source for older people not being more afraid of death is Cicirelli 2002. Fear of death in older adults. Journal of Gerontology, 57b.
That being said, I think that even though people do not become religious because they are afraid to die, religiosity decreases fear of death. The same study by Cicerelli indicates religiosity as a significant independent variable that predicts less fear of death.
I also have a cultural evolutionary scenario to support the hypothesis that fear of death is not the cause of religiosity, but that religions often develop attractive afterlives. In many cultures, the oldest afterlife beliefs are that people are in a shadowy, unattractive place (Sheol in ancient Judaism, Yomi in Shintoism, the Greek underworld).
See this post: http://www.cognitionandculture.net/home/blog/22-helen/2393-what-explains-foxhole-theism
Something is either true or not true. It is totally irrelevant what portion of the population believes what.
What people believe does not make beliefs true or false. What is at issue in the CCA is whether or not people's beliefs could constitute a form of evidence (even if prima facie, defeasible, weak etc) for the truth of a particular proposition.
Yeah, as others have stated, I think the crucial fact is that it matters how others have come to their beliefs. If a large majority believes in theism for vastly different reasons, or for reasons that we know not to be reliably truth-tracking (such as cultural transmission or cognitive biases), then it seems to me that the mere fact that these people are theists is not very good evidence for theism. Put another way, "theism is true" is but one of several possible explanations for the fact that many people seem to be theists, and it can be made more or less likely relative to other explanations depending on the other available evidence relevant to the mechanisms underlying the acquisition of religious belief.
For instance, I think we would have more reason to find the CCA compelling if we were to find that human psyches are not particularly predisposed to accept supernatural or intelligent agent-caused accounts of unexplained phenomena; that the kinds of religious belief subscribed to by people across time and space are relatively uniform in content and not (at least grossly) mutually contradictory; that cultural transmission is not a major influence on religious belief; or that most people hold religious beliefs on the basis of arguments that are robust to a reasonable amount of philosophical/empirical scrutiny. Of course I don't think any of those are anywhere close to being the case, so I'm not very inclined to change my evaluation of the epistemic probability of theism based on the worldwide prevalence of religious belief.
That said, I do like the example Alexander Pruss offers of being forced to choose between a path that everyone is following and a path that no one is following with the knowledge that one path leads to a good place and the other leads to a bad place. I think what this thought experiment demonstrates is that in the absence of any other evidence, the beliefs of a majority can be of at least some use in evaluating a proposition. Note however that the crucial assumption is that you have literally no other information to go on in making your decision; once we relax this assumption all the other factors I mentioned above come into play. (for instance, suppose that you discovered that the reason everyone is going the same way is that the first person who got to the fork picked a direction at random, and everyone else followed suit because they reasoned the same way the thought experiment asks us to. Then I think there would be no advantage to picking one direction over the other.) So while I think Mr. Pruss's argument demonstrates his point, I'm not sure that it has any practical relevance given the fact that we do have quite a bit of additional information about people's religious beliefs.
Anonymous:
"I think we would have more reason to find the CCA compelling if we were to find that human psyches are not particularly predisposed to accept supernatural or intelligent agent-caused accounts of unexplained phenomena"
Such a predisposition may itself be evidence of theism, of course, but we need to beware of double counting.
"for instance, suppose that you discovered that the reason everyone is going the same way is that the first person who got to the fork picked a direction at random, and everyone else followed suit because they reasoned the same way the thought experiment asks us to. Then I think there would be no advantage to picking one direction over the other."
Yes, but again only if we knew the case was precisely as described. In practice, we wouldn't know for sure that everyone (except the first person) was going to the left solely because they saw someone else doing it. In practice, we would suspect that, say, the negative fact that Smith did not receive strong evidence (say, from an inner voice) favoring going to the right is partly explanatory of his going to the left.
Let E be the claim that Smith did not receive strong evidence favoring going to the right. Then that Smith goes to the left is significant evidence for E. And E itself is weak evidence favoring going to the left. The fact that Smith goes to the left is thus* some evidence in favor of the hypothesis that going to the left is better.
Likewise, the fact that Smith believes in God, even if Smith's faith comes entirely from his parents, is significant evidence that Smith does not know a strong argument against the existence of God. And that Smith does not know a strong argument against the existence of God is itself weak evidence for the existence of God.
Footnote *: Of course, that A is evidence for B and B is evidence for C does not always imply that A is evidence for C. However, in the case at hand, we can make this move.
"And E itself is weak evidence favoring going to the left."
Doesn't this only work if Smith also does not receive strong evidence favoring going to the left (other than the fact that others are going to the left)? Remember that I've set this up so that the people going to the left are all doing it because they saw the others do it.
Suppose Smith asks a number of the people in front of him why they're going to the left and they all say "well, I saw everyone else doing it and I don't have any other information to go on, so it seems like the reasonable thing to do." Then indeed we could suppose that these people haven't received strong evidence in favor of going to the right, but we also can't suppose that they've received any evidence for going to the left above and beyond the fact that others are doing it. And since we know that the reason why others are doing it is because the first person picked a direction at random, it seems to me that this leaves us with no basis for the claim that going to the left is better. The first guy could just have easily gone to the right, and in that case everyone would be going to the right for the same reasons that they're now going to the left.
Similarly, if we suppose that Smith believes in God entirely because of his parents, it certainly is evidence that he doesn't know a strong argument against the existence of God, but it seems to be equally strong evidence that he doesn't know a strong argument for the existence of God (since if he did know one, then presumably his belief in God would be at least partially based on it. This assumes, of course, that Smith forms beliefs at least partly on the basis of arguments, but so does the claim that his belief in God is evidence that he doesn't have a strong argument against the existence of God).
Now you might say that the assumption that Smith's belief in God comes entirely from his parents is unrealistic (and you would probably be right), but then we're back to my original point, which is that it matters crucially for the CCA why people hold the beliefs that they do.
It seems to me that the bottom line is that the CCA is evidence for God's existence to the extent that the observed patterns of religious belief would be unlikely if the arguments in favor of God's existence were not stronger than the arguments against it.
"Then indeed we could suppose that these people haven't received strong evidence in favor of going to the right, but we also can't suppose that they've received any evidence for going to the left above and beyond the fact that others are doing it."
I think there is an asymmetry here. Let's suppose our evidence as to what evidence they have is that they sincerely say that they have no evidence and are just following their predecessors. Consider the hypotheses:
HL: These people have some decision-relevant evidence in favor of going left but are not aware of it under that description.
HR: These people have some decision-relevant evidence in favor of going right but are not aware of it under that description.
Now, I think the probabilities of HL and HR are small. But given our evidence, if the priors of HL and HR are the same, the posterior probability of HL is slightly higher than the posterior probability of HR. For HR decreases the probability of their going left (since we frequently act on decision-relevant evidence that we are not aware of as such), while HL increases it.
Consider two possible 'models'
1. God exists but some people cannot 'feel' him
2. God does not exist but most people think to 'experience' a God.
The first one is a much more 'simpler' model and thus more plausible.
Theodorus: 1 and 2 do not seem to be good comparisions because of the use of 'some' and 'feel' in 1 and 'most,' 'think' and 'experience'in 2. I think the wording makes 2 more complex, but unfairly so relative to 1.
If we use 3. God does not exists but some people 'feel' him., is 3 more complex or less complex then 1 or equal to 1?
John Alexander; Your 3 "God does not exist but some people 'feel' him" is logically incorrect so I think it is much more complex (as you need to explain what these people actually feel as it cannot be God)