You certainly know a story or a movie scene that involves hallucinations and you might even have experienced it yourself one way or another. Cult-status instances are given in movies like “Naked Lunch” or “Fear and Loathing in Las Vegas”:

Most people associate such experiences with psychedelic drugs or with psychotic states like schizophrenia. This association is so strong that it is part of the definition in the dictionary. But in fact, hallucinations can also be found in non-psychiatric and healthy individuals. To overcome the stigmatization of hallucinations, I will present some insights into this fascinating phenomenon and how its scientific investigation can lead to a better understanding of healthy and distorted perception.

It’s not real but it feels real.

Our perception of the world is a vivid and flexible experience that is hardly a direct representation of what is physically “out there”. Contextual interpretations and expectations about sensory information can lead to misperception of what is actually present. To understand the phenomenon of hallucination, we should start by distinguishing it from other forms of perceptual distortions. A hallucination is the perception of something that is not physically present but is experienced as vividly as any other form of sensory perception.

This immediately begs a comparison with dreams. But dreams are often blurry, they are fully immersive and usually don’t involve awareness. During hallucinations on the other hand, individuals are fully awake and the percept appears to be located in objective physical space. Contrary to imagined visual experiences, they are usually not voluntarily controlled by the individual. Hallucinated content appears suddenly and is usually observed passively. The percept can range from simple geometric figures or colours, smells and sounds to complex visual scenes, auditory conversations and even music that was (apparently) never listened to before.

Who experiences hallucinations?

Aside from the most prominent case of hearing voices and seeing intimidating creatures or insects, which can be symptomatic of schizophrenia and other psychoses, there is a variety of non-psychiatric disorders that are accompanied by visual hallucinatory experiences. Well-known cases are migraine, epilepsy, Parkinson’s disease, some types of dementia and the Charles-Bonnett-Syndrome. Vividly described by Oliver Sacks, the latter can occur when the visual sense is significantly reduced. Blindness that occurs later in life can lead to complex visual hallucination of people or scenes and, as recently described, can also include smells.

Similarly, a lack of sensory input can evoke hallucinations in healthy individuals. An impression of this effect might be experienced, for example, in sensory deprivation tanks: Floating in warm saline water in complete silence, surrounded by darkness may evoke spontaneous brain activity. Even though one hour of sensory deprivation might not be enough to have strong visualizations, in principle, everyone should be able to hallucinate. Some people could be simply more prone to it than others. Even higher consumption of coffee was found to predict hallucination proneness.

A large cross-national analysis of surveys from the world health organization with more than 30.000 respondents reports that about 5% of the general population has had a hallucinatory experience at least once in their lifetime. In older adults, as recently pointed out by a review article, a significant minority experiences hallucinations without having any psychological or cognitive disorders. Often, these experiences remain unreported. The stigmatization of hallucinations results in the fear of being judged as demented or mentally ill. Reasons for the occurrences of hallucinations in the elderly include decline in sensory functioning (similar mechanism as described above), social isolation, and poor sleep. The latter reliably induces hallucinations in humans in general.

Hearing voices is an infallible sign of psychosis, right?

Not at all. There is growing evidence for a significant percentage of the general population who have auditory verbal hallucinations (AVH) that are not related to psychotic criteria. When cognitive functioning was assessed in healthy people with AVH it seemed that the tendency to hallucinate in the verbal domain is related to a decrease in executive functioning leading to distractibility and problems with short term memory. Nevertheless, it is not clear if hallucinations are cause or consequence of such limited executive functioning. (If you hear voices all the time, how are you supposed to concentrate?)

Even though hallucinations are much more common than initially thought, their scientific investigation is not. Being primarily a subjective experience, the phenomenon is notoriously difficult to study. Researchers are starting to tackle the problem with simple but carefully designed studies combining  subjective reports with brain imaging methods. In a study published in Science this August, the authors made use of classical conditioning to induce auditory hallucinations in four different groups of participants. They investigated the differences in individuals with psychotic illnesses who hear voices daily (P+H+) and those who don’t hear voices (P+H-) as well as individuals without diagnosed illness who hear voices daily (P-H+) and those who don’t (P-H-).

While lying in the brain scanner, participants were presented a visual checkerboard stimulus paired with a tone. By pressing a button, participants were supposed to report if they heard the tone and the length of the button-press should indicate their confidence in hearing the tone. After multiple pairings, all participants tended to report a perception of the tone even in cases when the visual stimulus was presented alone. Interestingly, this effect was more pronounced in the two voice-hearing groups, regardless of diagnosed illness (both H+). The daily hallucinators also reported more confidence then the non-hallucinators. During these hallucinated auditory perceptions, the tone-responsive regions of the brain were more active than when participants correctly rejected hearing a tone. This was considered strong evidence for an actual perception of the hallucinated tone.

The researchers developed a computational model that assigned different weights to the participants beliefs and their sensory input. Fitting the model to the behavioural data revealed that hallucinators assign stronger weights to their prior beliefs about perceptual input (i.e. that a tone will be played together with the visual stimulus). Additionally, this bias correlated with activity in brain regions that are associated with expectation and beliefs.

As a next step, this perceptual model was used to distinguish psychotic participants (P+) from participants without the illness (P-) by means of their ability to update their beliefs (i.e. that the tone might not always be played). According to the model, psychotic participants update their belief suboptimally when changes in task contingencies occur. The belief-updating ability also correlated with brain activity in regions that are associated with sensory and motor predictions.

Hallucinating your reality

The researchers consider their experiment evidence for the “strong-prior hypothesis of hallucinations”. The world is a changing place and the human brain makes sense of this uncertainty by integrating prior experience with new incoming sensory information. (In this article, I describe how similar mechanisms of sensory integration can influence your sense of self in out-of-body experiences and other illusions.) According to the “strong-prior hypothesis”, some people assign stronger weights to their prior belief about the world, which, in turn, influences their perception. In extreme cases, this might even lead to the perception of something that is not present at all. One aspect of the psychotic brain might be that this prior belief is not updated accordingly.

“Your brain hallucinates your conscious reality”. This phrase by Anil Seth describes that the entire spectrum of our conscious experience is generated internally by the brain. Sometimes it relies more on sensory information and sometimes it generates conscious content just by itself. By using all information that is available, the brain produces its best guess on what reality is.