Hallucinations as a disorder of perceptual inference


Leandro Castelluccio

Schizophrenia and dopamine

Finding or seeing patterns that do not really exist is a problem of schizophrenia. At the same time, finding unusual patterns in reality is a key aspect of our creative capacity. There seem to be a link between both, and dopamine might be the underlining factor between them. A greater propensity to see novel patterns could underlie creative people and people with schizophrenic disorder. It has been shown that high creative abilities are somewhat more common in people who have mental illnesses in the family, creativity being linked to a slightly higher risk of schizophrenia and also of bipolar disorder. Certain psychological traits, such as the ability to make unusual and extravagant associations are also shared by schizophrenics and highly creative non-schizophrenic people alike. Finding novel or unusual patters is in turn related to dopamine. For example, when studying certain receptors in the brain, it has been demonstrated that the dopamine system in healthy and highly creative people is similar in some aspects to that observed in people with schizophrenia. It is observed that highly creative people who obtained good results in divergent thinking tests had a lower density of D2 dopaminergic receptors in the thalamus than the less creative people. It is known that schizophrenics have a low-density level of D2 receptors in this part of the brain, which suggests a link between mental illness and creativity.It is thought that lower D2 receptors in the thalamus result in a lower degree of signal filtering and therefore a greater flow of information from the thalamus, which could be a possible mechanism behind the ability of both healthy and creative people to see numerous unusual connections in a problem-solving situation in the same sense as people with schizophrenia can see strange associations that do not match reality (see in link).

We must also take into account that schizophrenia is related to the dopaminergic pathways in its positive symptoms. The most frequent positive symptoms reported by patients are hallucinations (usually auditory), irrational delusions-beliefs irreducible to logic and alterations in thinking (disorganization, incoherence).

The drugs indicated in this disorder are those of the group of neuroleptics, which act on the neurotransmitter dopamine. Second-generation or atypical neuroleptics differ from first-generation neuroleptics because they are more specific for the mesolimbic pathway and less so for nigrostriatal, which is involved in extrapyramidal side effects. They also block the serotonergic 5HT2 receptors increasing the action of serotonin in the limbic system, which is associated with the clinical improvement of the negative symptoms of schizophrenia.

Difficulties with this kind of explanations of positive symptoms

According to Fletcher and Frith (2009), explanations such as “hallucinations are caused by overactive dopamine receptors” are unsatisfactory because they leave an explanatory gap between the mental and the physical in the sense this unusual dopamine activity  does not tell us how a voice is hallucinated or how a person has an illogical belief? Additionally, following these authors, the evidence that deluded patients have difficulty with standard logical-reasoning tasks is limited, where the symptoms that are typically reported suggest something more subtle than a general problem with reasoning.  And although hallucinations and delusions can be found in many neurological and psychiatric disorders, there is a subset of these symptoms that is more specifically related to a diagnosis of schizophrenia. This subset of symptoms includes hallucinations and delusions that are particularly related to a sense of being passive, of being subject to the control of an external force or agent.

According to Fletcher and Frith (2009), cognitive theories about the positive symptoms of schizophrenia, have tended to treat perception and belief formation as distinct processes. However, recent advances in computational neuroscience seem to indicate that the unusual perceptual experiences of patients and their sometimes bizarre beliefs are part of the same core abnormality, which would be a disturbance in error-dependent updating of inferences and beliefs about the world. Fletcher and Frith (2009)suggest that it is possible to understand these symptoms in terms of a disturbed hierarchical Bayesianframework, without recourse to separate considerations of experience and belief.

A new Bayesian approach

Hallucinations are false perceptions, such as patients hearing people talking about them or hearing their thoughts spoken aloud. Delusions are persistent bizarre or irrational beliefs that are not easily understood in terms of an individual’s social or cultural background. For example, patients may believe that other people can hear their thoughts or that the government is monitoring their every action (Fletcher & Frith, 2009). 

According to these authors, a common mechanism, involving minimization of prediction error, may underlie perception and inference, and that a disruption in this mechanism may cause both abnormal perceptions (hallucinations) and abnormal beliefs (delusions). 

According to new models of perceptual inference and predictive coding, the primary cause of positive symptoms is an abnormal sensory experience (a hallucination). Delusions follow as a secondary consequence of attempts to understand the anomalous sensory experience. For example, if patients can hear their own thoughts being spoken aloud (hallucination), it would seem logical to conclude that other people can also hear them (delusion) (Fletcher & Frith, 2009). 

As we mentioned before,it is thought that lower D2 receptors in the thalamus result in a lower degree of signal filtering and therefore a greater flow of information from the thalamus. This could be linked to the fact thattheories of abnormal perception assume that the fundamental problem underlying positive symptoms is the loss of the distinction between relevant and irrelevant stimuli.At the lower, sensorimotor levels the disturbance could be seen as subtle changes in perceptual and motor function, both of which have been noted in schizophrenia. In terms of experience, stimuli might feel unusual, important and salient, leading to a difficulty in allocating attention to appropriate aspects of the environment. This might be experienced as an enhancement of background sounds or sights that should be irrelevant, a phenomenon that is clearly described by patients in the early stages of the illness (Fletcher & Frith, 2009).But this is just one aspect of this new approach.

As Fletcher and Frith (2009) mention, probabilistic reasoning is associated with a Bayesian approach to the study of belief formation. Within this framework, a belief is the subjective probability that some proposition about the world is true. This probability is continually updated by new evidence. However, abnormal belief formation occurs when beliefs are not updated appropriately on the basis of new evidence. Such abnormalities in the integration of new evidence into beliefs have been observed in deluded patients. It has been shown that the synchronization of activity in brain areas that are associated with self-produced speech is attenuated in people with schizophrenia, and that this strongly predicts their likelihood of experiencing hallucinations. If a person cannot correctly track sensations or voices as his own (giving rise to an hallucination), and use prediction-errors to accommodate for this, this would quickly evolve into false believes about the world. As Fletcher and Frith (2009) point out, predictability is a useful marker for internally generated actions, but its significance extends to external stimuli, which can also be more or less predictable since we can learn when and where they will arise and with which other stimuli they are likely to be associated. 

Thus, for the world not to remain confusing, elemental and disjointed, each perception of a stimulus must update the internal record of its probability (the beliefs about the world), so when a new stimulus-outcome pairing occurs, prior occurrences of elements of that pair (the stimulus and the outcome) must be evoked if one is to have any meaningful idea of whether it is truly an association or simply a chance co-occurrence (Fletcher & Frith, 2009). As these authors mention, how one experiences something at a basic sensory level is dependent on one’s knowledge of it (expressed in terms of its predictability), where each experience is affected by what one believes. Furthermore, the extent to which one updates what one believes is affected by how that experience adds to it. This is the insight that was captured by Bayes’ theorem.

However, there is little neurocognitive evidence about the brain responding to prediction errors in general with regards to different sensory modalities, except in the field of reward, which is linked to dopamine and which is at the same time linked to schizophrenia, as we mentioned before.

Following Fletcher and Frith (2009),key mesolimbic dopaminergic regions are highly responsive to reward and prediction error. Furthermore, reward and prediction error-related behaviors are sensitive to dopaminergic perturbations. For example, a drug that up regulated dopamine function in healthy participants strengthened error-dependent reward learning, whereas a dopamine-blocking drug reduced such learning. There is also evidence that this system is disrupted in schizophrenia: the ventral striatum (a key component of the mesolimbic dopamine circuit) was shown to be under-active in response to rewarding experiences in patients. In addition, abnormal reward-based prediction-error activity has been observed in midbrain and target regions in people with schizophrenia (Fletcher & Frith, 2009). 

A problem with this is that patients with schizophrenia also often have negative symptoms, which accentuate over time. The negative symptoms of schizophrenia are dullness, anhedonia (inability to experience pleasure), apathy (lack of initiative or motivation) and impoverishment in language. It could be argued that the lack of observed response to the reward is related rather to the aspects of negative symptomatology of the disorder, rather than to its positive symptoms, and that error-dependent reward learning problems for example, are not specific problems of the ability to predict and update based on errors, but a problem of the level of reward response that people with schizophrenia naturally experience.

So the model proposes a new, fresh way perhaps, of explaining positive symptoms in schizophrenia, but there are many aspects not covered by this model, such as negative symptoms of the disorder, or why positive symptoms arise more frequently around 20 years of age and not before. 


Da Silva, F. R. (2009). Manual de psicofarmacología clínica. Prensa Médica Latinoamericana.

Dopamine system in highly creative people similar to that seen in schizophrenics, study finds. (2010). Retrieved from https://www.sciencedaily.com/releases/2010/05/100518064610.htm

Fletcher, P. C., & Frith, C. D. (2009). Perceiving is believing: a Bayesian approach to explaining the positive symptoms of schizophrenia. Nature Reviews Neuroscience10(1), 48.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s