The evidence for credulism
The format will be converted to that of this site, and the content revised, shortly.
Credible Evidence for Credulism
First published here 4 September 2013
An archive of subsequent changes is available via the 'View history' tab above-right.
You believe every claim that you hear or read.
You probably think that this is obviously untrue - indeed, your likely disbelief of this claim may itself seem to disprove it. However, research seems to show that our mere comprehension of a claim involves believing it - something first proposed by the seventeenth-century philosopher Baruch Spinoza.
It seems obvious, from experience, that our mere comprehension of a claim doesn’t involve believing it, that belief occurs, if at all, subsequent to comprehension, upon assessing the claim to be true. However, in my article ‘How Belief Works’ I present an argument which seems to show, on the basis of logic alone, that merely comprehending a claim must indeed involve believing it. Here, I present the experimental evidence.
But in case you’re so skeptical that you’re currently wondering whether it’s worth reading any further, I should first address the above objection that this theory seems to be obviously contrary to our experience. I address this point in detail in ‘How Belief Works’, but briefly:
First, this theory doesn’t state that we indefinitely believe every claim that we comprehend, but simply that we believe claims upon comprehending them, which doesn't exclude the possibility of us subsequently ceasing to believe them. And, given how fast the brain functions, our belief of claim X, upon comprehending it, may be replaced by a contrary belief - which may have also been our belief prior to comprehending X - so soon after the formation of belief X that we‘ve no recollection of it.
For example, upon comprehending the claim ‘Paris is the capital of Germany’, we may seem to disbelieve it without at any point believing it. However, according to this theory, our comprehension of the claim actually involved believing it, but that belief only lasted the fraction of a second that it took us to remember that Paris is actually the capital of France, not Germany, and we therefore have no recollection of our brief belief of the above claim.
The argument in ‘How Belief Works’ doesn’t just concern claims entering our mind upon comprehending them, via an analysis of speech sounds, text, etc., but also to claims entering our mind upon being generated by our mind as we’re thinking. That is, not only are we always gullible when we’re comprehending claims, we also always jump to conclusions when we’re thinking - although, as with comprehension, such beliefs may only last a fraction of a second. However, the currently available experimental evidence solely concerns claims entering our mind upon comprehending them.
I call this theory, that we believe every claim that enters our mind, whether it does so upon being comprehended or upon being generated by our mind, ‘credulism’, given its claim of our complete credulity.
2 The evidence
The psychologist Daniel Gilbert - author of the international bestseller Stumbling on Happiness - wrote an academic article called ‘How Mental Systems Believe’, which supports the theory that comprehending a claim involves believing it. He presents a different theoretical argument to that presented in ‘How Belief Works’, but also documents experimental evidence.
2.1 Our innate credulity versus our learned incredulity
Gilbert points-out that this theory of comprehension accords with an aspect of our psychological development that should be familiar to everyone, but which has also been confirmed experimentally: there’s an inverse relation between age and credulity. He writes:
- Children are especially credulous, especially gullible, especially prone toward acceptance and belief - as if they accepted as effortlessly as they comprehended but had yet to master the intricacies of doubt.
If we always believe, at least initially, whatever we comprehend, then we have an innate, and constant, bias towards believing what we’ve comprehended. But, as we get older, that bias will at least be increasingly counteracted by our increasing knowledge, which can challenge new claims, and which includes our increasing awareness, from experience, that people sometimes lie, or wrongly believe, however genuine, or certain, they may seem, and that our own certainty can also be misplaced. Gilbert also argues that our innate credulity, and learned incredulity, is also reflected in our linguistic development: research has shown that the ability to express disbelief arises much later than the ability to express belief.
It might be objected that our youthful ignorance will itself lead to a strong bias towards believing, which decreases with age. That is, if it’s only as we get older that new claims are increasingly challenged by our increasing knowledge, then, even under the current theory of belief formation - with belief and disbelief being alternate outcomes of an assessment subsequent to comprehension - we must start-off, in early childhood, believing every claim that we’ve comprehended, with this credulity decreasing with age. Therefore, the inverse relation between age and credulity doesn’t necessarily imply an innate, and constant, bias towards believing, that we’re able to increasingly counteract with age, because it could simply be a product of our youthful ignorance.
However, if our belief of a claim is the product of reasoning subsequent to our comprehension of it, then we must have a reason to believe it. And not having a reason to disbelieve a claim isn’t the same thing as having a reason to believe it - it’s simply a reason to suspend disbelief until a reason to believe, or disbelieve, is found. For example, not holding a belief about the population of Mongolia isn’t, in itself, a reason to believe someone’s claim about the population of Mongolia. Of course, people, especially children, may still wrongly think that the absence of a reason to disbelieve a claim is, in itself, a reason to believe it. But it could also instead be wrongly thought that the absence of a reason to believe a claim is, in itself, a reason to disbelieve it - and there’s no reason to think that either wrong thinking would be more likely than the other.
Regarding an unawareness of people’s capacity to lie, or be wrong, again, this wouldn’t, in itself, constitute an awareness of a reason to believe a comprehended claim - we’d also have to explicitly believe that people are always truthful, and never wrong. And we’d in turn need to have reason to reach this false belief. But it seems very unlikely that even young children would consistently initially think that they had such a reason, and then hold that false belief for some time. Indeed, we surely become aware of our own capacity to lie, or wrongly believe, almost immediately after developing the capacity to comprehend, assess, generate, and communicate claims, even if our ability to then maintain awareness of the capacity of the human mind to lie, or wrongly believe, is initially low.
So, under the current theory of belief formation, our youthful ignorance should actually reduce our ability to believe comprehended claims as much as our ability to disbelieve them, and should therefore produce a tendency to suspend belief.
2.2 Forced belief
Gilbert also points-out that the theory that comprehending a claim involves believing it accords with the use of sleep-deprivation as an indoctrination technique. This technique is based on the assumption that it’s easier to get someone to believe something if you first deplete their mental resources. Gilbert quotes one political prisoner’s account of the effect of sleep-deprivation:
- You are annihilated, exhausted, you can’t control yourself, or remember what you said two minutes before. You feel that all is lost. From that moment the judge is the real master of you. You accept anything he says. [emphasis added]
This effect of depleted mental resources on belief formation has also been replicated in the lab. Of course, depriving experimental subjects of sleep to the extent that they can barely think isn’t allowed these days. However, what experimenters can do is ask subjects to listen to claims while performing an unrelated task, such as watching an amusing silent film after being told that they’ll be asked questions about its content afterwards. Given that we’ve a finite amount of mental resources, this method is a humane way of significantly reducing the amount of such resources subjects can allocate to what they’re hearing. And several studies have found that when subjects listen to a short talk arguing in favour of a view that’s contrary to their own, there tends to be a greater shift in their view towards the other view if their mental resources are significantly reduced while they’re listening to the talk.
It’s been suggested that the reason why mental exhaustion, or distraction, leads to such changes in belief is that the resulting significant reduction in available mental resources significantly reduces the person’s ability to perform the assessment, subsequent to comprehension, required for disbelief. However, if the current theory of belief formation is correct, and belief is an alternative outcome of that assessment, then mental exhaustion, or distraction, should equally reduce a person’s ability to believe what they’re hearing, and should therefore tend to cause them to suspend belief.
But the fact that mental exhaustion, or distraction, causes such changes in belief obviously accords with credulism: the resulting significant reduction in available mental resources significantly reduces the person’s ability to use their knowledge to disbelieve what they believe upon comprehending. In the extreme case of sleep-deprivation, people are effectively returned to their childhood level of credulity.
2.3 Credulity without credulism
Although the experimental observations mentioned so far - the inverse relation between age and credulity, and the increased credulity under depleted mental resources - accord with credulism, and seem incompatible with the current theory of belief formation, they don’t actually demonstrate that comprehending a claim involves believing it, because there’s another possibility: our belief and disbelief of a claim are alternate outcomes of an assessment subsequent to comprehending the claim, but, contrary to the current theory, the assessment is, for some reason, strongly biased towards belief, with that bias being increasingly counteracted with age as described with respect to credulism. In the case of the experimental subjects listening, under depleted mental resources, to a talk arguing in favour of a view that’s contrary to their own, their increased credulity can be explained by them tending to at least have sufficient resources to reach the low threshold for belief, via the arguments given in the talk, but not the higher threshold for disbelief.
However, this alternative theory doesn’t seem to be able to explain the reported increased credulity when sleep-deprived targets of indoctrination are simply presented with claims that are contrary to what they believe, rather than also being presented with alleged reasons to accept those claims. Again, if our belief of a claim is the product of reasoning subsequent to our comprehension of it, then we must have a reason to believe it - even if that reasoning is strongly biased towards belief.
Also, Gilbert presents other experimental observations which seem to disprove this alternative theory, and prove that comprehending a claim involves believing it.
2.4 Believing what we’ve reason not to
One striking experimental result has been replicated in several studies. In these studies, subjects simply listened, without distraction, to an audio recording of a stranger reading aloud a position statement, such as ‘I think that abortion should be illegal’. Even though the subjects knew that the stranger was obeying an instruction to read aloud the statement - which therefore provided a motive for the verbal statement that didn’t involve the speaker believing it - the subjects tended to be subsequently found to believe that the reader held the position in the statement that they’d read.
According to the current theory of belief formation, the subjects’ belief of the statement would have depended on an assessment subsequent to their comprehension of it. And, again, if our belief of a claim is the product of reasoning subsequent to our comprehension of it, then we must have a reason to believe it. But even if such an assessment was strongly biased towards believing, the subjects would have had no good reason to conclude that the stranger held the position in the statement that they’d been instructed to read aloud. Of course, the absence of a good reason to believe the statement doesn’t mean that a subject couldn’t have thought of a reason which they thought was a good reason, but which was actually a poor reason - such as ‘They sound like someone who would believe that’. But it seems reasonable to assume, especially given that such an assessment wouldn’t have been conducted under depleted mental resources, that most subjects would have correctly assessed that the facts of the situation meant that there wasn’t a good reason to believe the statement, and that they should therefore suspend belief.
However, according to credulism, the subjects’ belief of the statement would have been simply dependent on their comprehension of it. It might nevertheless be thought that even credulism can’t account for the experimental result, given how fast the brain functions. That is, the subjects’ awareness of the facts of the situation would have meant that their belief of the statement, upon comprehending it, would have been replaced by a suspension of belief well before the experimenters had time to determine the subjects’ belief about the stranger’s position. However, this would have tended not to happen, for the following reason.
If someone believes that it’s raining outside, then their belief could cease either because they observe that it isn’t raining outside, which reverses their belief, or because they realise that it’s so long since they looked out of a window, and observed that it was raining outside, that it could have since stopped, and they therefore suspend their belief about whether it’s raining outside. In the first case, the content of the belief is incompatible with a fact, whereas, in the second, it’s not that the content of the belief is incompatible with a fact - they haven’t observed that it isn’t raining outside - but that the belief itself lacks a sound justification, because that it’s so long since they observed the weather outside.
Regarding the above experiment, although the facts of the situation meant that the subjects’ belief of the statement, upon hearing it, lacked a sound justification, those facts weren’t actually incompatible with the content of the belief. That is, just because the stranger was obeying an instruction to read the statement aloud didn’t, in itself, mean that they didn’t hold the position in the statement. Therefore, the subjects’ awareness, subsequent to hearing/believing the statement, of the facts of the situation wouldn’t, in itself, have led to that belief ceasing - they would have also needed to then reason that those facts meant that there was no reason to believe the statement, and that they should therefore suspend belief. And while such reasoning is quite basic, and there wasn’t anything deterring them from so reasoning - such as having to attend to some other task - there was equally no reason why they would have been more likely to do so than thinking about numerous other possible things, such as what they thought of the position itself, or whether the experiment would be coming to end soon, or who the stranger’s voice reminded them of, or what time they’d arranged to meet a friend later that day. Within credulism, not only is it not necessary to establish a justification for believing something before believing it, it isn’t necessary to establish such a justification upon believing it. Moreover, the possibility of these numerous other trains of thought also means that the subjects may not even have considered, subsequent to hearing the statement, the facts of the situation. All of this would therefore explain the tendency for the subjects’ belief of the statement, upon comprehending it, to endure.
Gilbert also presents an even more striking experimental result. Subjects were told that they would be presented with a series of pairs of suicide notes, and that, for each pair, they would be asked to identify which was real and which was fake. They were also told that, after each decision, they would be immediately given feedback on their performance, but that this feedback wouldn’t actually be determined by their performance, but would be completely arbitrary. Nevertheless, when, after they’d considered 25 pairs of notes, they were asked to estimate how well they’d done, there was a significant correlation between their estimates and the feedback that they'd received.
If belief of the feedback was the product of an assessment subsequent to comprehending it, then, even if that assessment was strongly biased towards believing, at least most, if not all, of the subjects would have taken into account the advice about the arbitrariness of the feedback, and the feedback therefore wouldn’t have been believed. But if comprehending a claim involves believing it, then the prior advice wouldn’t have prevented the subjects from believing the feedback upon comprehending it.
As with the previous experiment, it might be thought that the subjects’ belief of the feedback would nevertheless have then been almost immediately replaced by a suspension of belief - this time, given their awareness of the prior advice. But, again, this would have tended not to happen, for the same reason. Although the facts of the situation meant that the subjects’ belief of the feedback, upon hearing it, lacked a sound justification, those facts weren’t actually incompatible with the content of the belief. That is, just because the feedback was arbitrary didn’t, in itself, mean that when, for example, the feedback consisted of ‘correct’, the subject hadn’t been correct in their conclusion about the current pair of notes. Therefore, the subjects’ awareness, subsequent to hearing each feedback, of the prior advice wouldn’t, in itself, have led to their belief of the feedback ceasing - they would have also needed to then reason that that advice meant that they should suspend belief with respect to the feedback just heard. And while such reasoning is quite basic, and there wasn’t anything deterring them from so reasoning - such as having to attend to some other task - there were also numerous other possible things to think about, such as the content of the current pair of notes, or the reasoning which led to their decision, or what it would feel like to write a suicide note, and what they’d put in it, or what they were going to do after the experiment. Again, within credulism, not only is it not necessary to establish a justification for believing something before believing it, it isn’t necessary to establish such a justification upon believing it. Moreover, again, the possibility of these numerous other possible trains of thought means that the subjects may not even have considered, subsequent to hearing the feedback, the prior advice. And if we also consider that, in order for the feedback to have not impacted on the subject’s assessment, they would have had to consider the prior advice, and then its implications for believing the feedback, upon hearing each feedback after each of their 25 decisions, then it seems that credulism can indeed account for the correlation between the feedback and the subjects’ estimates of their performances.
3 Support for Spinoza
So Spinoza seems to have been right: merely comprehending a claim seems to involve believing it. It’s of course ironic that his claim has been largely unbelieved for over three hundred years. However, the above research, and the earlier explanation of why this theory of comprehension isn’t actually contrary to our experience, together mean that we’ve no good reason to disbelieve this theory, subsequent to comprehending/believing it.
Share this article
If you liked this article, then please consider supporting my writing by sharing this article with others.
- Shortened URL: http://goo.gl/q4ebLq
Discuss this article
Constructive feedback is always gratefully received. Please post any objections, questions, or other comments, about this article - however minor - in the Trying to Think Google Group.
- Farnell, D., 2013, ‘How Belief Works’, Think, volume 12, issue 35, pages 39-60.
- Gilbert, D., 1991, ‘How Mental Systems Believe’, American Psychologist, volume 46, number 2, pages 107-119.