On the failure of the Outsider Test for Faith
October 1, 2010 § 19 Comments
John Loftus is well-known defender of Atheism, especially on the internet (which has led him to more than a few spats). He’s written one book and edited another, arguing against Christianity, particularly the ‘fundamentalist’ kind.
The centerpiece of his intellectual assault is the ‘Outsider Test for Faith’: the idea that, if believers one religion treated it with the same scepticism as they treated other religions, they would lose their faith. In other words, religious faith is driven by distortions about being ‘within’ a faith group, and taking an ‘outsider’ point of view will correct it.
I’m not a believer, but I’m also not convinced that the OTF is the rhetorical silver bullet it is made out to be. I hope to clarify and augment the OTF to avoid some of the more common criticisms, and hopefully cut through some of the confusion between Loftus and his detractors. In the final reckoning, though, I will show the OTF isn’t a significantly persuasive force for Atheism.
What’s the test?
What exactly is the outsider test? “to test their own adopted religious faith from the perspective of an outsider with the same level of skepticism they use to evaluate other religious faiths” isn’t entirely clear. This lack of clarity – both from Loftus and his critics – makes the argument very hard to dissect indeed.
Loftus has offered a number of developments about the OTF. It started here in 2006:
So let me propose something I call The Outsider Test: If you were born in Saudi Arabia, you would be a Muslim right now, say it isn’t so? That is a cold hard fact. Dare you deny it? Since this is so, or at least 99% so, then the proper method to evaluate your religious beliefs is with a healthy measure of skepticism. Test your beliefs as if you were an outsider to the faith you are evaluating. If your faith stands up under muster, then you can have your faith. If not, abandon it, for any God who requires you to believe correctly when we have this extremely strong tendency to believe what we were born into, surely should make the correct faith pass the outsider test. If your faith cannot do this, then the God of your faith is not worthy of being worshipped.
It’s most modern statement is this: Loftus (2010).
. Rational people in distinct geographical locations around the globe overwhelmingly adopt and defend a wide diversity of religious faiths due to their upbringing and cul-tural heritage. This is the religious diversity thesis.
Consequently, it seems very likely that adopting one’s religious faith is not merely a matter of independent rational judgment but is causally dependent on cultural condi-tions to an overwhelming degree. This is the religious dependency thesis.
Hence the odds are highly likely that any given adopted religious faith is false.
So the best way to test one’s adopted religious faith is from the perspective of an out-sider with the same level of skepticism used to evaluate other religious faiths. This expresses the OTF.
So this is the newest version (a book is in the works), and the contours of it follow prior developments. There’s a big problem, though, pointed out by Paul Manata and Victor Reppert: it isn’t a valid argument! There’s no way  follows from , nor  from &, nor  from . (Another worry is that Loftus often comes across as doctrinaire: it’s just said that if we’re rational or fair or unbiased in evaluating religious beliefs we turn to Atheism, but what distortion believers have, or how the outsider test is meant to correct for it, remains unclear).
In practice, however, the ‘outsider test for faith’ takes on more than a few different meanings, in part because of ambiguity surrounding ‘outsiderness.’ It is often implied that to be rational, believer’s need to take an ‘outsider’ perspective. To subtract away evidence only available ‘inside’ a given religion. Insofar as these implications are made, they are wrong. Suppose you are floating on the wreckage of a downed ship, and you hear the BBC world service say your ship was lost with all hands. This would be very good reason for an ‘outsider’ – that is, someone outside your immediate ‘epistemic locality’ to believe you are dead. Yet of course you, clinging to the wreckage have very good reason to believe you are alive. The fact you have no means to convince an outsider of this fact doesn’t mean you should disregard this sort of evidence. In short, on this subtraction story, failing the OTF is the reasonable thing to do.1 What you should be committed to is some sort of epistemic parity commitment – that a reasonable person, finding themselves in your particular epistemic situation, would form the same beliefs that you do. But everyone thinks that.
Perhaps we can do Loftus an intellectual charity. We can see the sort of argument he is driving at, even if he hasn’t expressed it well. What he wants to say is that the particular ‘internal’ or ‘privileged’ experiences that religious believers take at evidence are really epistemic delusions, that trick believers into what they believe. Let us try to produce an argument which is logically valid, and uses the sort of demographic data Loftus alludes to to try and convince someone that that is so.
Here’s my attempt. Consider it a renovated version of the ‘start’ of the OTF. Call it the ‘demographic undercutting defeater’ (DUD).
5) The best explanation of the demography of religious belief is ‘cultural inertia’. [Premise]
6) If the best explanation for a demography of beliefs is X, most of those beliefs are driven by X [premise]
7) Most religious beliefs can are driven by ‘cultural inertia’ [from 5 and 6]
8) Beliefs that are formed through ‘cultural inertia’ are unwarranted. [Premise]
9) Most religious beliefs are unwarranted. [from 7 and 8]
This captures the work of the first three premises, and follows Loftus’s approach: one infers from the demographics of religious belief that the best explanation for this pattern is ‘cultural inertia’: that is, of ‘undue influence’ by ones social/cultural milieu to lead one to be prejudiced in favour of the prevailing religious view of your environment. Most (but not all) religious belief can be explained this way. Yet beliefs driven by cultural inertia are plainly unwarranted. They fail any epistemological metric you care to name (PF, Reliable process, Tracking). So most religious beliefs are unwarranted. It follows that a religious belief you hold is probably unwarranted.
This seems also fairly resilient to prior criticisms of the OTF.  seems strong: it seems to have any potential theist counter-offer (eg. A daemonic influence theory) beat on the explanatory virtues front.  and  are uncontroversial, and  and  follow robustly.2
We could give a much simpler argument against Theism, following Malitzen (2006):
10) The religious pattern of earth is better explained on Atheism than on Theism
11) Atheism is confirmed over Theism by some increment. [10 & Prime principle of confirmation]
This also seems solid. If religions were nothing more than particular social phenomena, then we’d expect them to show the geographically and culturally delimited patterns they do. Theistic explanations of the same have a much harder time.
Yet this isn’t what Loftus has in mind. Instead of pointing to evidence for Atheism, he wants to point to a particular cognitive distortion he thinks affects Theists, and offer a corrective so they can realize their distorted cognitive lens and correct it – that is, the outsider test. We can cash out a similar plea from DUD: if most religious beliefs are formed by cultural inertia, and are consequently unwarranted, decent cognizers want to avoid being one of the most who have beliefs distorted in this way. So subtracting out unwarranted cultural inertia and seeing where that leads may be a good strategy. If this ‘outsider’ to cultural inertia finds in favour of your religion, you’ve ‘passed’ the OTF, or my renovated version thereof.
So that’s the argument, which I hope is pretty faithful to what Loftus has in mind when he presents the OTF. This version also makes clear why the genetic fallacy doesn’t apply: with warrant and justification and so on, the ‘genetics’ of a belief is exactly what we’re interested in. A belief, even if it is true, may have a poor epistemic pedigree. In the real world, being confronted that our beliefs are unwarranted, the defense that they could be true regardless is lame. So if the charge of the OTF (or DUD) sticks, this is bad news. Does it?
What should we be testing?
Loftus has religious belief in mind in the OTF, yet it is clear these sorts of arguments have a broader scope. Any beliefs that show considerable cultural plasiticity are susceptible to these sorts of demographic defeaters. As Hays notes, attempts to religion alone should fall under scrutiny seem highly question-begging and tendentious: they amount to stating that Christianity isn’t warranted. Reppert suggests the Outsider Test for Faith should really be the Outsider Test for Belief. This seems about right: for any topic on which opinions are highly culturally plastic, you should wonder whether it’s really just cultural inertia driving you to the attitude supported by your cultural background.
Is this problematic? Yes, but not in the way Loftus’s critics charge. Manata suggests (among others) two counter-examples to the OTF: a) taste, and b) external world skepticism (Manata 2010). Neither seem lethal counter-examples: a) is unproblematic – we’re happy to be fairly ecunumenical about tastes, or accept that our taste really is culturally plastic (I like chips, but if I grew up elsewhere I might hate them, so what?) I don’t think b) has much purchase either: beliefs uniformly held by the human population (like external world realism) don’t suffer from the warrant objection posed in DUD.3
What we need is a fair way of telling which beliefs are susceptible to OTF-esque objections. Just saying it should apply to religion seems self-serving and (as far as I can see) poorly argued. Happily, I think we can provide this. We can look at what we’d expect the demography of belief to be if it was arrived at by a warranted process.
Scientific beliefs neatly pass this test. Affirmation of relativity (or, more pointedly, evolutionary biology) correlates with stuff like expertise, educational level, and so on: stuff we think tends to be warranted. Scientists from all different cultural backgrounds end up believing similar sorts of things about (for example) human anatomy. So we can be confident (barring some conspiracy theory about the scientific academy) that these beliefs about human anatomy were arrived at in a warranted manner, and haven’t been formed through cultural inertia.
Religious beliefs plainly aren’t like this. People by and large do follow the religious beliefs of their prevailing cultural situation.4 If people really were ‘following the evidence’, we’d expect the pattern of belief to follow the pattern of evidence to a degree dependent on the epistemic aptitude of the population in question – so people would hold ‘random’ religious beliefs if there was little between them, or there would be a distinctive trend of the better-informed heading to one religion if the evidence base was slanted in its favour. In fact, religious beliefs are such that the best predictors are things like the time and place where you live, the religion of your parents, and so on5: this makes little sense on an ‘informed inquiry’ model, but a lot of sense on a ‘cultural inertia’ model.
The problem is there are other beliefs like this too: moral beliefs. These too show a similar level of cultural plasticity. Go back a couple of hundred years, or displace me a thousand miles or so, and I’d likely be racist, sexist, and homophobic (although, of course, I flatter myself in thinking I would be a rebel). So, via parallel arguments, we see that most moral beliefs aren’t warranted – and ours probably aren’t too. We have strong convictions against thinking our moral convictions unwarranted. Yet, obviously, if we have problems with people persisting in Christianity despite this ‘demographic defeater’, surely we commit the same epistemic sin just believing our moral beliefs are warranted despite the probability of cultural inertia?
One ‘out’ is to accept moral irrealism – but it would be a very ‘hard-core’ version of irrealism. Accepting our moral beliefs are just the shifting of the cultural sands means that any hope of normativity in ethics is lost. Regardless of what we ‘should’ believe, odds are we’ll probably believe whatever our prevailing cultural norms so determine.6 Another is to shrug and say “all our community of speakers will believe the same sorts of things about not killing people – so we can’t be sincerely challenged here”. Certain stripes of non-realists might be happy at this, but not all. Most of us want to defend our deep moral convictions instead of trusting them to be ‘taken as read’ by all likely interlocutors. Besides, controversial moral beliefs aren’t like this, even in our own place and time.
It seems the ‘best’ out is to assert that our particular environment gives us privileged access to true (if not realist: better, apter) moral beliefs. Yet such an approach seems obviously high-handed. Perhaps we can bulk up the case by tracking developments in moral attitudes and the hope that there’s something that set us apart – that our principles are obviously better defended, or that there’s a conversion assymmetry to ‘our point of view’. Yet note how this discussion parallels the one around the OTF and religious believers can make exactly these sorts of moves. The demographic defeater isn’t irrebuttable, so we can cast around for justifying reasons for our culturally plastic belief that weather the undercutting force here.
Look into my mind…
Let’s suppose though there is a distinction you can draw between the two cases, so DUDs show ‘selective’ toxicity to religious beliefs alone. What then?
Something vital that has been missed so far. Our demographic data only can show generalities. It doesn’t show that every single instance of someone believing a religion is driven by cultural intertia, although it suggests that most are. Likewise, although it suggests that for any person with religious beliefs, that the reason for those beliefs is probably cultural inertia, it isn’t certain that they are. There could be a minority that form religious beliefs through honest, even-handed enquiry.
Why care? Because there’s no way for us (or Loftus) to know for sure a particular person is one of the unwarranted majority or the reasonable minority. They could be one of the few people who have conducted a fairly free and somewhat fair inquiry, and that lead them to reaffirm the beliefs of the prevailing cultural environment. Although some believers may ‘not think about this stuff’ and find confrontation with the OTF genuinely novel, most of Loftus’s detractors (and pretty much every single apologist on the internet) all affirm that they have thought about things carefully without making any obvious question-begging or similar – that, further, they can present cases that should convince a reasonable outsider to agree with them.
The rejoinder is that these people are deceiving themselves about how objective they’re being. Because it is suspect that all these people who confess being even-handed and objective just-so-happen to end up reaffirming their prior beliefs far more so than would be expected from really even-handed and honest enquirers. That seems fair – from this, it might be fair to assume that many people who say they are honest and even-handed enquirers have subconsciously stacked the deck already.7 Again though, this wouldn’t prove there was no one who wasn’t an honest and even-handed enquirer who really did find this confirmed their prior views were on the money. Further, for particular sorts of believers (eg. Converts from other religions, re-converts), it seems hard to give a ‘cultural’ story that isn’t contrived8 – in other words, this is evidence that counts in favour of them likely being honest enquirers.
Basic commitment to good faith and charity should lead one to accept that your opponent is correct when he affirms these things about himself – of course, you may suspect that most such people are mistaken about themselves, but you take him at his word (besides, it’s not like he’s going to take your word over his with regards to the internal functioning of his own faculties). To do otherwise is to seem perilously close to the evangelical who locates all disbelief as a character defect of being blinded by sin.9
Suppose you confront a believer with the demographics of belief in some form of OTF or DUD or whatever. I think the main responses can be grouped in two main ways.
- Some believers, who have pretty much defaulted on to the religious belief of their upbringing may well be surprised. It may jolt them in to considering whether they do have good reasons for their religious convictions besides ‘going with the flow’ of their environment. It may prompt deconversion, either by recognition that they have no good cause to continue their religious belief, or through leading them to enquiry that leads them to Atheism.
- Other believers (like those you find on the internet) will be pretty literate about philosophy and other matters. Some may offer a number of arguments that they think both demonstrate their religious beliefs are the correct ones, and offer these arguments to others to persuade them of the same. In short, they think there’s a favourable public case for their religious beliefs, even if they also affirm (like Craig) that it would be unnecessary to motivate their religious beliefs due to the overwhelming private case too.
Although there are some type-1 responders, most of the responders are type-2. Yet it is to the type-2s the OTF has barely any relevance whatsoever. These people genuinely think that the public case is in their favour. Even if it could be shown that their own personal revelation is decisively undercut, what does it matter? The conversation evolves like this.
“Let us accept that my religious experiences and other ‘first person’ evidence are undercut by the demographic data you present. Let us discount these, and take an ‘outsider test’. Well, I have this bit of natural theology (e.g. Argument from Reason, argument from design, etc.) that doesn’t rely on my religious experiences. So, as an outsider10, I’d be persuaded by this argument, and thus become a believer. So my beliefs pass the OTF.”
“Not so fast. I reckon although you say you’ve discounted your religious beliefs, I think they’re colouring your evaluation of these conclusions. You wouldn’t find this bit of natural theology persuasive if you weren’t a believer.”
“But this was the very argument that convinced me to become a Theist! How can you possibly be so confident that I reason in this defective way? It doesn’t seem so to me, and I’m going to take my word over yours as to the state of my own mind.”
This doctrinaire (and logically rude) assertion about how people evaluate arguments without even considering the arguments has to go. The next moves must be to discuss the public case the believer finds so convincing, and to evaluate whether it really is. If you want to persuade the believer to unshackle themselves, this is surely more likely to be successful (unless, as is often charged, the OTF is something for non-believers to backslap one another with).
If you really think the believer is hopelessly deluded about the merit of these arguments you should either a) not talk about it or b) refute it for the benefit of any other parties to the conversation. Just asserting irrationality doesn’t cut it. (Again, compare to someone who cites romans 1 at you after hearing your carefully developed problem of evil).
Shutting the door on the Outsider Test for Faith
There are both superficial and deep problems with the OTF as Loftus has it. I hope my renovated OTF solves the superficial issues, but the deep ones remain regardless. There are arguments around the demography of religious belief that are good (again, Malitzen (2006)). The OTF, alas, over-reaches itself. It tries to evolve from promising data an entire epistemic mechanism to translate believers out of their mistake. In fact, it seems often to be used as a roundabout way of asserting believers are irrational because they haven’t ‘taken’ this test (a test which is often vague and, somewhat conveniently, the marker of ‘success’ is agreeing that your opponent is right!)
The superficial problems are that a) the OTF as presented isn’t a valid argument, and b) it is often deployed with a tendentious presumption that any sort of ‘internal’ evidence from a given religion must, on principle, be rejected. I think my treatment irons out those creases: the DUD seems valid, and further gives a principled reason to be suspicious of ‘internal’ evidences like religious experience – that they seem to show a high degree of cultural determination that suggests they are unreliable, don’t track, etc.
Yet the deeper problems remain. We might be willing to say (if we’re ‘friendly’ in the Rowean sense) that religious believers aren’t driven by culture, but simply have non-overlapping, non-intersubjective yet reasonably persuasive grounds for their diverse beliefs. Even if we don’t accept that, we find that similar problems apply to moral beliefs (and probably the justification of moral beliefs, and the sort of moves we want to make to avoid the teeth of that issue can be mirrored by the believer to get out of danger too. Worst of all, it seems dialectically misguided – many believers do consider themselves to have good reason to believe. Assuming they are wrong and dismissing them as irrational without even addressing their reasons is neither irenic nor persuasive.
Loftus, J. (2010) The outsider test for faith revisited. In: Loftus, J. (ed) (2010) The Christian Delusion: why faith fails New York: Promethius Press. Ch. 4.
Malitzen, S. (2006) Divine hiddenness and the demographics of theism. Religious studies 42 177-191.
Manata, P. (2010) A review of Chapter 4. In: Chan, P., Engwer, J., Hays, S., and Manata, P. (2010) The Christian Delusion. Available at: http://www.calvindude.com/ebooks/InfidelDelusion.pdf
1Here you may object: “but there are ways for our liferaft-person to convince an outsider. They just need to see them.” Perhaps, but a) the illustration applies elsewhere – suppose a powerful first person experience of (say) the existence of God, or an intuition that gay couples aren’t worse than straight couples. These fail inter-subjectivity in a similar manner, but no one thinks that makes them epistemically outlawed.
Loftus may say there’s something fishy about these sorts of first person sources of evidence that means we shouldn’t take them as seriously as we do. This will be explored later. The point is that a doctrinaire “discount all evidence you couldn’t convey to an outsider” isn’t reasonable by its own lights. It is the success of the OTF explored later that determines to what degree these sorts of evidences are impeached.
2Of course, proper functionalists that this sort of de jure objection just begs the de facto question. I don’t think that move works here. The foregoing can easily be cashed out as a de facto argument against God. Bear with the epistemic slant for now.
3It may pose problems for Loftus’s original argument. However, I often can’t work out what Loftus has in mind when he refers to the OTF. So far as I can tell, though, his version is sufficiently flawed that a reductio to external world scepticism is the least of that argument’s worries.
4Perhaps this applies to atheism too. My hunch is the argument that sets Atheism apart is not all that bad, but hardly powerful: that people from all different places renounce their religious beliefs without adopting other ones, and there’s something different about not accepting any religious beliefs as opposed to accepting one-or-the-other, in the same way there’s something different about not playing a sport or not having a particular taste in music. I avoid exploring the ‘what about Atheism?’ objection, not because I think it lacks merit, but I think there’s a far stronger one.
5There are disputes about how people really are converting to Christianity from Islam, or that there’s a conversion assymmetry so that well-informed Christians deconvert more than well-informed atheists convert, but I’m not convinced by any side here.
6In other words, if you’re a fan of forms of moral non-realism that preserve some normative sense (e.g. expressivism) you are stuffed. Needless to say, it is highly toxic to folk moral practice.
7But maybe not. With people who seem to be fulfilling their epistemic obligations in excelsis, like professional philosophers and so on, maybe it is rather the case that – recalling our shipwreck case from earlier, different religious believers have been supplied by their experiential background non-overlapping sets of ‘first person’ evidence that testimony can’t convey persuasively to an outside party. This might be reading Dostoyevsky, hearing the recitation of the Quran, seeing the horrendous evil of your daughter dying of leukaemia, and so on. Even if we discover that these sorts of experiences are decisive in forming our religious convictions (which I strongly suspect they are), I don’t think that makes those convictions unreasonable, even if we recognise that, if we had experienced things differently, our beliefs too would differ.
8This is perhaps why de/conversion narratives are taken so seriously. That someone ‘beat’ their cultural inertia and chose another path is rhetorical gold to the converted-to, and painful to the converted-from. Hence all the wars over recontextualizing conversoin narratives.
9It’s called logical rudeness. Generally, people tend to be quite smug when confronted by logical rudeness and they aren’t being logically rude themselves. They (somewhat ironically, meta-rudely) take the offering of a logically rude story to explain away their disagreement a sign of weakness – for they are happy to assume standard epistemic norms for their opponent.
10‘Outsider’ is perhaps a misleading term. I’d prefer something Bayesian like subtraction or background knowledge. Note here I’m assuming we retain (‘stay inside’) standard epistemic norms and doxastic practice. Some have suggested presupesque responses like ‘once we remove faith, what hope is there for reason?’ or Manata’s ‘defeater deflector’. Perhaps. I don’t think anything so controversial is needed to neuter DUDs like the OTF. However, if you find that an argument like “if God did not exist, epistemic norms would carry no force, and I believe they carry force”, that argument can ‘pass’ the OTF.