Perceptual experience is a distinctly privileged way of knowing about the world. Not only is perceptual experience ultimately the bridge between mind and world, but it also trumps other ways of knowing. When another way of knowing about the world – inference, introspection, memory, and testimony – and experience disagree with one another, experience will be kept and the other will be dropped unless we have strong reason to believe that what we're experiencing contains an illusion or a hallucination. So, if you were to tell me that my sister is in Melbourne, and later that day I saw her walking across the street from where I am here in Adelaide, I would immediately drop the belief based on testimony that she is in Melbourne. However, an alternative situation may be this: I know that my sister has a doppelgänger who lives here in Adelaide. So in this situation, if I happen to believe that you're a reliable source of information, I'd probably believe instead that I'm actually seeing my sister's doppelgänger.
The moral is this: experience, privileged as it is, is still judged against what we already happen to know – as are the other sources of knowledge. When a proposition arrived at by inference or testimony disagrees with something I already know, I'm going to subject that proposition to much closer scrutiny than I otherwise would. Furthermore, experience, privileged as it is, always involves interpretation. In the first case, where I see my sister, and the second case, where I see my sister's doppelgänger, provide me with identical data. Each of the perceptual experiences are indistinguishable from my first-person perspective. The fact that interpretation is involved in perceptual experience is what explains how it's possible to come to different conclusions from identical perceptual experiences.
So, how are we to understand “interpretation” in the context of perceptual experience? It's certainly not anything like conscious deliberation, otherwise its presence would be salient to us (and it's not at all), and further, experience would be much more plastic than it actually is, in that it would be affected by interpretation in a much more thoroughgoing way. So our background knowledge influences our perceptual experiences in a way that is automatic and unconscious. Should this consideration lead to scepticism about the reliability of our perceptual faculties giving us objective knowledge of the external world? I think the answer is clearly not.
To give an example from Jerry Fodor's excellent (and hilarious!) paper “Observation Reconsidered” (1984), consider the Müller-Lyer illusion:
We experience the bottom line, with the inward facing arrowheads, as being longer than the top line, with the outward facing arrowheads. We know perfectly well that both lines are of equal length, but that knowledge doesn't penetrate our perception. What does penetrate our perception is a set of rules or heuristics that we bring to perception after a lifetime of being raised in environments with ninety degree angles that relate to us in systematic ways (people raised in hunter-gatherer societies, for example, are less susceptible to the illusion). Corners of rooms are always going outwards from our perspective, and corners of tables are always coming towards us in our perspective. The two-dimensional nature of the Müller-Lyer illusion is a counterexample to the heuristic that fits the three-dimensional world almost perfectly well. So why don't the facts of interpretation and heuristics raise sceptical worries about perceptual experience? Because the interpretation and heuristics present is perceptual experience is extremely reliable. When we look at the Müller-Lyer illusion we misperceive what's actually there. But, importantly, we don't have the experience of a sunset, or a waterfall, or a TED talk. Our misperception is an uncanny facsimile of what's actually there.
So, perception of the physical world is extremely reliable. But we humans are primarily social beings. How are we to evaluate our perception of the social world? Social beings are, of course, much more complex than mere physical beings like corners and tables, so we need to come to the social world with much more complicated heuristics. These are heuristics that we humans have arrived at via two stages: from having evolved in socially complex environments (as well as physically complex environments); and from being raised and enculturated in socially rich environments. This gives us a grasp of folk psychology, which is a framework we use for understanding and predicting each other's behaviour.
Our perceptual ability in the social world is remarkable considering the complexity of what it is we're doing when we interact with one another (this is so effortless for most people that it's often taken for granted). Just by perceiving a few subtle changes in a person's facial expression, we can see exactly how they feel about news they're hearing, the film they're watching, or the food they're eating. We can read off the quite complicated contents of a person's mind from a very impoverished sample of their behaviour. Furthermore, we have a remarkable ability to coordinate with one another, like when thousands of scientists collaborate at CERN, and just when hundreds of motorists can share a road without crashing into each other. In either case, with realising it, we're perceiving and processing all sorts of information that is contained in each other's mind.
Of course, our remarkable perceptually abilities stemming from folk psychology can be a little “too good,” as it were. This can lead us to misperceive and attribute psychological properties all over the place when there actually are none to be found. One example is the overattribution of agency. We humans are raised in environments where most changes (or most changes that matter to us) are brought about by conscious agents. So it's an essential part of living in a human environment that we are able to attribute agency and keep track of it, and this is done in a remarkably reliable way. However, a necessary byproduct of this ability is that agency will be overattributed. We attribute conscious agency to animals, which may be well and good as long as those attributions are not overly anthropomorphic. But equally, we attribute agency to the weather, to desktop computers, to cars that won't start, and so on. The same story can be told for emotions. It's crucial for living in a socially complex environment that we can correctly attribute emotions to ourselves and the people around. Again, a byproduct of having this ability is that it gets overextended, and hence that's why we can sometimes experience the weather as being hostile towards us, or cars and computers as angry at us when they're playing up. Salient examples of this sort of overextension are in things like Cydonia (the face on Mars) and the Shroud of Turin. In both cases we can't help but see a human face, but this is only because the heuristics that shape our experience don't account for “accidental” faces.
This leads me to consider distinctively religious experience. Insofar as religious experience is a form of perceptual experience, as it often will be – after all, perception is the bridge between mind and world, and religious entities are supposed to have an existence independent of my own mind – what can be said about the nature of religious experience? We've noted that perceptual experience is distinctively privileged, and yet involves interpretation and heuristics which don't undermine perception, but instead seem to do the opposite. One difference between perceptual experience in general and religious experience in particular is that the latter seems to be a wholly private affair, whereas the former can be subjected to the many perspectives of public scrutiny. The privateness religious experience is something that's often used by the religious to get themselves off the hook of rational criticism. Arguments for God, such as the design argument or the Kalām cosmological argument can be criticised from the public perspective – a perspective sharable by everyone – as being fallacious. Arguments for God based on religious experience, on the other hand, would ipso facto seem to be immune to public scrutiny, because the experience happened in one particular perspective and not in the usual case of many perspectives. The upshot then being that theists who've had religious experiences have a privileged form of justification that's not available to others.
I think drawing this sort of conclusion from a religious experience is wrong, based on what I call the opacity of religious experience. I don't deny that such experiences exist – they obviously do – but I do want to put pressure on what such experiences are about. Perceptual experience as such is always about something; it always has some propositional content or other. In the case of perceiving things in the physical world, our experiences are about physical objects and their properties. In perception we are aware of tables and chairs and the like. In the case of illusion, like the Müller-Lyer illusion, we are aware of the lines themselves, but we just get some of their properties wrong (having different lengths, for example). In social perception, we are aware of social objects, like other minds or emotional states. In both of these cases, the objects of perception are more or less transparent to us. Our perceptions are about the lines in the Müller-Lyer illusion and about the contents of our best friend's mind.
Religious experiences are quite different to this because of their elusive and ethereal nature. Such experiences can have a very wide range of interpretation, and hence, coming to a religious experience with even slightly different heuristics to support the experience and make sense of it can drastically influence the experience itself. I may be going through an incredibly warm and blissful state that I perceive as the presence of God, but I may have an identical, indistinguishable experience while on psychedelics that I may perceive as being at home in the universe, or being one with the universe. This is just a more dramatic example of drawing radically different conclusions from absolutely identical experiences, and it all depends on what heuristics and background knowledge we bring to the table. However, the difference in the case of religious experience, unlike experience of the physical and social world, is that interpretation of the experience seems to be entirely open-ended. And it seems to me that the only explanation for this open-endedness is that what religious experiences are about is wholly a matter of interpretation, and therefore ascriptions of God and other entities in religious experience is entirely unwarranted.