New research suggests that paranoia and excessive teleological thinking—two forms of distorted beliefs about intention—are associated with specific errors in visual perception. The findings, published in Communications Psychology, indicate that individuals with high levels of paranoia are more likely to misperceive non-threatening movements as chasing, while those with strong teleological thinking tendencies struggle to accurately identify agents in social scenarios. The results suggest a link between these cognitive distortions and basic visual processes.
People frequently attribute intentions to others, even in ambiguous situations. This tendency extends beyond human interaction, as individuals can perceive intention in the movements of simple shapes. The classic example is the Heider-Simmel animation, in which geometric figures appear to chase, evade, and interact with one another, even though their movements are entirely pre-programmed.
Previous research has suggested that the rapid detection of agency and intention is an essential function of human vision, allowing people to react quickly to potential threats. However, when this ability becomes exaggerated or distorted, it may contribute to mental health conditions such as paranoia, which involves an excessive belief that others have harmful intentions. Similarly, excessive teleological thinking—the tendency to attribute purpose to events that lack clear intentionality—has been linked to various cognitive biases and belief systems, including conspiracy thinking.
The researchers sought to determine whether these higher-level cognitive distortions could be traced back to fundamental perceptual processes. Specifically, they examined whether individuals with paranoia and excess teleological thinking were more likely to misinterpret visual cues related to agency and intention, leading to what they termed “social hallucinations.”
“My lab studies psychosis – delusions and hallucinations – which are the profound departures from consensus reality that characterize serious mental illnesses like schizophrenia, but which also exist in milder forms in the broader, general, population,” said study author Philip R. Corlett, an associate professor of psychiatry and psychology, a member of the Wu Tsai Institute and the director of the Belief, Learning, and Memory Lab.
“We try to understand the mechanisms that make people think that their neighbors are spying on them, for example. Furthermore, we want to know why they might draw that particular conclusion versus a different one. Placing these symptoms in the perceptual or cognitive domains has implications for how we understand and treat psychotic symptoms and also what we think about how minds and brains work.”
The researchers conducted a series of four studies with a total of 623 participants. They used an experimental paradigm in which participants watched displays of moving discs and had to determine whether one disc (the “wolf”) was chasing another (the “sheep”). This allowed the researchers to measure perceptual errors related to the detection of chasing behavior.
In the first two studies, participants simply judged whether a chase was occurring. The researchers manipulated the movement patterns of the discs to create ambiguous situations in which no actual chase was present. False alarms—instances where participants incorrectly perceived chasing when none was occurring—were used as indicators of social hallucinations.
In the third and fourth studies, participants had to identify which disc was the “wolf” and which was the “sheep.” This allowed the researchers to assess whether paranoia and teleological thinking were associated with specific impairments in role attribution. Participants also rated their confidence in their decisions, which provided insight into the extent to which they believed in their misperceptions.
Alongside the behavioral task, participants completed questionnaires assessing paranoia, teleological thinking, and self-reported hallucinatory experiences. These measures helped the researchers examine how perceptual errors correlated with broader cognitive tendencies.
The results showed distinct patterns of perceptual errors associated with paranoia and teleological thinking. Individuals with high levels of paranoia were more likely to falsely perceive chasing when none was present. This suggests that paranoia may involve an exaggerated sensitivity to potential threats, even at the level of basic visual processing. Notably, these participants were particularly poor at identifying the “sheep” in the chasing displays, which may reflect difficulty in recognizing non-threatening agents.
In contrast, individuals with high levels of teleological thinking were more likely to struggle with identifying the “wolf”—the agent that was supposedly pursuing the other disc. This suggests that excessive teleological thinking may involve a different kind of misperception, possibly related to difficulties in distinguishing between intentional and unintentional movement.
Both paranoia and teleological thinking were linked to a tendency to make high-confidence errors. Participants who exhibited these cognitive tendencies were often certain about their incorrect judgments, suggesting that their perceptual distortions were deeply ingrained. Additionally, these errors correlated with self-reported hallucinatory experiences, reinforcing the idea that the misattribution of agency in vision may be linked to broader tendencies toward perceptual distortions.
“People with unusual beliefs about other people and the sorts of things that cause events to happen in the world may actually see the world differently,” Corlett told PsyPost. “They see intentional actions where others don’t and they have these experiences with high confidence – which is why we term them social hallucinations.”
Although the findings provide strong evidence that paranoia and excessive teleological thinking are linked to specific perceptual biases, the study has some limitations. One major question is whether these distortions originate in low-level visual processing or are instead the result of top-down influences, where pre-existing beliefs shape perception. Future research could explore this question using neuroimaging techniques to examine the neural basis of these perceptual errors.
Another limitation is that the study focused on non-clinical populations. While the results suggest potential links between perceptual distortions and psychiatric symptoms, further research is needed to determine whether these findings apply to individuals with diagnosed psychotic disorders like schizophrenia.
“We need to test these ideas in people with schizophrenia – who have more severe hallucinations- to see if they can be practically useful in the clinic,” Corlett said. “We’d like to develop this game into an eye test for psychosis. It could be useful in identifying people at risk, or people whose symptoms might be worsening.”
The study, “Paranoid and teleological thinking give rise to distinct social hallucinations in vision,” was authored by Santiago Castiello, Joan Danielle K. Ongchoco, Benjamin van Buren, Brian J. Scholl, and Philip R. Corlett.
