Short answer: the difference lies in decision threshold, not in the eye
When looking at the same ambiguous image, people with mystical beliefs tend to report 'there is a face' more often than skeptics. But on closer analysis, researchers found that true sensitivity to visual signals was nearly identical between the two groups. The difference is the response threshold: mystical believers set a lower decision criterion, meaning that just a slight face-like cue is enough for them to say 'yes'. In other words, they do not have a sharper eye, but a more lenient criterion when deciding.
This conclusion comes from research by Tapani Riekki and colleagues published in 2013 in Applied Cognitive Psychology3. The authors had 47 participants complete two face-detection tasks in blurred landscape images. Mystical believers and religious individuals recognized face-like regions more often, but they also made more false alarms. When applying signal detection theory to separate sensitivity from response threshold, the researchers found that true sensitivity did not differ from skeptics. This is evidence level B: the results are suggestive but the sample is small and not widely replicated, so caution is needed when generalizing.
When told there might be a face, everyone sees faces in pure noise
If the difference is not in the eye, where is it? An experiment by Jiangang Liu and colleagues published in 2014 in Cortex answered this quite clearly2. The researchers showed participants pure noise images, which contained no real shapes, but told them that 50% of the images contained faces or letters. Result: participants reported seeing faces in 34% of images and letters in 38% of images, even though neither actually existed2. This shows that merely suggesting 'there might be a face' leads the brain to automatically interpret meaningless light and dark patches as faces.
More notably, when monitoring brain activity with fMRI, the researchers found that the right fusiform face area, the region specialized for processing real faces, also activated strongly when participants 'saw' faces in noise images2. This means the brain does not clearly distinguish real from illusory faces at the sensory processing level. When signals from the visual cortex travel upward, higher processing areas send feedback signals downward, causing the face recognition area to interpret ambiguous signals as faces2. This is a top-down mechanism, meaning expectations and beliefs shape what we see.
Larger samples at fairs and online: mystical believers report seeing more faces
Michiel van Elk, a psychologist at the University of Amsterdam, conducted two field experiments at a spiritual fair in the Netherlands, published in 2015 in PLOS ONE1. In the first experiment, fair visitors were asked to classify images with varying noise levels as 'face' or 'house'. Results showed that people with strong mystical beliefs tended to report 'face' more often than skeptics, but they also made more false alarm errors1. Van Elk concluded that mystical belief is associated with a more lenient response threshold for ambiguous stimuli, not with better signal detection ability1.
The second experiment in the same study tested whether belief in conspiracy theories relates to visual attention. Results showed that conspiracy believers tended to be more influenced by local information when processing global images, suggesting that different types of beliefs may be linked to different cognitive biases1.
Signal detection theory: a four-cell table explaining why 'seeing more' does not mean 'being better'
To understand why reporting more faces does not equate to seeing better, one needs to know the basic principle of signal detection theory. When deciding 'is there a face or not' in an ambiguous image, each answer falls into one of four cells:
| Actually has a face | Actually no face | |
|---|---|---|
| Participant says 'yes' | Hit | False alarm |
| Participant says 'no' | Miss | Correct rejection |
True sensitivity is measured by the ability to distinguish 'actually present' from 'actually absent', i.e., the hit rate relative to the false alarm rate. Response threshold is the participant's readiness to say 'yes', regardless of whether the signal is strong or weak. A person with a low threshold will say 'yes' in more cases, leading to both more hits and more false alarms. A person with a high threshold will be more cautious, missing more but making fewer false alarms.
In van Elk's study, mystical believers had a higher hit rate but a correspondingly higher false alarm rate, so overall sensitivity did not differ1. Van Elk suggested that skeptics may rely on more diagnostic or detailed features to distinguish faces from non-face objects, while mystical believers use less stringent criteria1.
Why is the brain so sensitive to faces?
The ability to recognize faces is one of the most important visual skills of humans, because faces carry essential social information. Liu and colleagues noted that face pareidolia, seeing faces in clouds, toast, or inanimate objects, has been recorded for centuries2. This suggests that the human visual system is highly tuned to detect faces, possibly because of their social importance2.
This mechanism has evolutionary significance: in ancestral environments, missing a face (especially a stranger's or a predator's) could be far more dangerous than mistaking an inanimate object for a face. Thus, the brain tends to lean toward saying 'yes' when signals are ambiguous. This tendency is not an individual error but a common human trait.
Beliefs shape perception: from faces to illusory patterns
Research by Jan-Willem van Prooijen and colleagues published in 2018 in the European Journal of Social Psychology extended this question to a broader domain. Across five studies, the authors found that people who tend to see patterns in random coin toss sequences or abstract paintings also tend to believe more in conspiracy theories and supernatural phenomena. Notably, when researchers actively primed participants to perceive patterns, their belief in conspiracy theories and the supernatural also increased.
This finding suggests that the link between illusory pattern perception and supernatural belief is not only correlational but may be causal in one direction. However, note that these results come from studies conducted mainly in Western countries; no similar research has been done on Vietnamese people. The general mechanism of seeing patterns where none exist is detailed in the article Why humans see patterns where none exist. The current article focuses on belief and response threshold, without repeating that general mechanism.
'Apparition' photos spreading on social media: suggestion and expectation effects
This phenomenon explains why photos claimed to show 'apparitions' spread strongly on social media. When someone posts a photo with a suggestion that it contains a shadow or face, the next viewer already has an expectation. As Liu and colleagues demonstrated, merely suggesting there might be a face in a noise image leads most participants to report seeing one2. This effect does not depend on whether the viewer believes in the supernatural.
When an ambiguous photo is shared with a story, viewers no longer see it as a meaningless image but search for faces in it. And because the brain is already sensitive to faces, the result is that many will 'see' what the poster wants them to see. This is the effect of expectation on perception, a phenomenon well documented in cognitive psychology.
Beliefs can produce real outcomes: distinguishing effect from accurate prediction
One point needs clarification: there are cases where beliefs genuinely lead to different outcomes in reality, but not because a spiritual system 'predicts correctly'. For example, if a person believes they can communicate better with others through spiritual practice, they may actually pay more attention to nonverbal cues, thereby interacting more effectively. In this case, the belief changed behavior, and behavior changed the outcome. This is a real effect, but it does not prove that the spiritual system predicts events accurately. It is important to distinguish these two to avoid both denying the effect and attributing it to supernatural ability.
Where evidence is still lacking
All existing studies were conducted in Western countries, mainly Finland, the Netherlands, and North America. No research has been done on Vietnamese people to test whether results replicate in a different cultural context. Riekki and colleagues' sample had only 47 people, a fairly small number for firm conclusions3. Van Elk's results were collected at a spiritual fair, where participants may not represent the general population1. Van Prooijen's studies on the link between illusory pattern perception and supernatural belief also relied mainly on student and online samples.
It is important not to extrapolate results from these studies to Vietnamese people without local data. Culture, religion, and ways of expressing spiritual belief in Vietnam may differ significantly from Western countries, and those differences may affect how people respond in similar experiments.
Summary table of main studies
| Study | Method | Sample size / scope | Main result |
|---|---|---|---|
| Riekki et al. (2013)3 | Two face detection tasks in blurred landscape images | 47 people | Mystical believers report more faces, more false alarms; sensitivity unchanged; more lenient response threshold |
| Liu et al. (2014)2 | Viewing pure noise images, told faces or letters present; fMRI measured | Not stated in abstract | Participants reported faces in 34% and letters in 38% of images; right fusiform face area active when 'seeing' faces |
| van Elk (2015)1 | Classifying noise images as face or house at spiritual fair; visual attention test | Spiritual fair visitors | Mystical believers report more faces; more lenient response threshold; conspiracy believers show local attention bias |
| van Prooijen et al. (2018) | Five studies on pattern perception in random sequences and abstract art | Various samples | Those seeing illusory patterns believe more in conspiracy and supernatural; priming pattern perception increases belief |
What really makes the difference?
In summary, existing evidence suggests that the difference between believers and non-believers is not in visual ability. Both groups have the same biological machinery for face detection. The difference is the decision criterion: when is it certain enough to say 'yes'. Mystical believers tend to accept a lower level of evidence before concluding that a face or supernatural phenomenon exists. This does not mean they see worse or are more easily fooled, but that they assign different weights to two types of error: missing a possible phenomenon and falsely reporting one that does not exist.
This conclusion also aligns with Liu and colleagues' finding that when primed, ordinary people also see faces in pure noise images2. In other words, we all have the potential to see faces where none exist, and each person's decision threshold can change depending on context, expectations, and beliefs. Understanding this mechanism helps us view 'seeing ghosts in photos' or 'seeing faces in objects' as a common feature of human perception, rather than as a sign of special ability or gullibility.
