My name is Magdalena Antrobus, I am a PhD student working on Project PERFECT, researching psychological and epistemic benefits of depression. Together with Lisa Bortolotti I wrote a paper entitled Depressive Delusions, exploring the nature of delusions in severe forms of depression as well as the process of their formation. Here we present a summary of the article, which was published in 2016 in the Filosofia Unisinos journal.
It is common to define delusions as implausible beliefs that are held with conviction but for which there is little empirical support. The vast majority of delusions appearing in severe depression are mood-congruent, which means that their content matches the mood experienced by the person (Hales and Yudofsky, 2003). Common themes of depressive delusions are persecution, guilt, punishment, personal inadequacy, or disease, with half of the affected people experiencing delusions with more than one theme.
Stanghellini and Raballo (2015) point to several differences between schizophrenic and depressive delusions. People affected by schizophrenia often describe the adoption of the delusion as a discovery, such as the discovery of the true meaning of life, or of a new purpose for humanity (Stanghellini and Raballo, 2015, p. 173), although delusions can and often do incorporate aspects of the person’s everyday reality and past experience. Delusions that emerge in depression – on the other hand - confirm self-related information that is already known and familiar. Delusions of guilt – for example - may validate a feeling of guilt and confirm the person’s conviction that she has done something wrong.
It is common to define delusions as implausible beliefs that are held with conviction but for which there is little empirical support. The vast majority of delusions appearing in severe depression are mood-congruent, which means that their content matches the mood experienced by the person (Hales and Yudofsky, 2003). Common themes of depressive delusions are persecution, guilt, punishment, personal inadequacy, or disease, with half of the affected people experiencing delusions with more than one theme.
Stanghellini and Raballo (2015) point to several differences between schizophrenic and depressive delusions. People affected by schizophrenia often describe the adoption of the delusion as a discovery, such as the discovery of the true meaning of life, or of a new purpose for humanity (Stanghellini and Raballo, 2015, p. 173), although delusions can and often do incorporate aspects of the person’s everyday reality and past experience. Delusions that emerge in depression – on the other hand - confirm self-related information that is already known and familiar. Delusions of guilt – for example - may validate a feeling of guilt and confirm the person’s conviction that she has done something wrong.
People with severe depression acquire increasingly negative beliefs about themselves because the process by which they acquire self-related information is disrupted. The balance between assimilation and accommodation is compromised in that people with depression become increasingly less able to use accommodation as a way of processing new self-related information. Instead, they distort the content of newly acquired information in order to assimilate it successfully and making it consistent with their previous beliefs. The expansion of negative self-schemata is reflected in the judgments that people with depression express about themselves: ‘I am worthless’, ‘I have done nothing good in my life’, ‘I deserve to be punished’, and so on.
According to the cognitive model, preserving consistency is a priority for human cognition (Festinger, 1957). People have a general tendency to restore coherence between their views, and between their views and the views of others, when this is compromised. If a person finds herself in a state of extreme cognitive dissonance, she needs to validate her prior beliefs and offset the cognitive dissonance emerging from those beliefs and newly acquired information. Mood-congruent delusions in depression play this key role. In particular, they may offer validation for one’s own intensively experienced guilt, shame, hopelessness, and dismay.
The claim that one is being spied upon in one’s own home – for example – matches the belief that one is not trustworthy and should be monitored at all times. Undoubtedly such delusions come with psychological costs. The delusional content is weaved with unpleasant, sometimes terrifying events, such as being watched, followed, or threatened. But, by reducing dissonance, delusions also offer psychological relief from dissonance–related anxiety. The delusion–induced anxiety that replaces the dissonance-related anxiety has the advantage of reinstating consistency: the negative beliefs about the self are confirmed.
Our hypothesis is compatible with the prediction-error theory of delusion formation (e.g. Fletcher and Frith, 2009; Corlett et al., 2007) which has already been applied to delusions in schizophrenia. According to this model delusions are formed in response to aberrant prediction-error signals, those signals that indicate a mismatch between expectation and actual experience (Miyazono et al., 2014). A prediction error happens when new incoming information does not match the person’s existing representations (schemata) and, therefore, cannot be successfully integrated in the person’s model of the world (assimilated). It indicates that the internal model of the world from which the prediction is derived is incorrect and needs to be updated. By updating the model in such a way as to minimize prediction errors, the person gains a better understanding of the world.
In the case of depressive delusions, the process differs as people err on the side of conservatism as opposed to revisionism. Instead of changing their representation of the world to match the unusual experience, validating new experience at the expense of previously acquired beliefs, people with depressive delusions preserve their representation of themselves despite the new information that conflicts with it, validating the existing self-schemata at the expense of the new information that is reinterpreted and distorted to fit the schema. When this happens, an adaptation process has to occur for the system to regain cognitive balance.
Here we suggest that mood-congruent, motivated delusions in schizophrenia are epistemically innocent. Here are the two conditions for the epistemic innocence of delusions: 1. Epistemic benefit: The adoption of the delusional hypothesis confers a significant epistemic benefit to a given agent at a given time, 2. No alternatives: Alternative hypotheses that would confer the same benefit are not available to that agent at that time.
Some qualifications are in order. First, what counts as an epistemic benefit may vary. One might say that adopting a delusion is epistemically beneficial if it contributes to the acquisition or retention of true beliefs, if it promotes the agent’s intellectual virtues, or if it is something an agent should be praised and not blamed for. One’s commitments in epistemology will affect the way in which the epistemic benefits are identified and described. In the case of depressive delusions, the main benefit seems to be the preservation of a coherent self-representation.
Second, different notions and degrees of unavailability can explain the failure to adopt a less epistemically costly hypothesis. This spectrum of possibilities reflects the nature of the limitations that the agent experiences in the relevant context, ranging from standard reasoning limitations affecting all human agents to deficits of perception, inference, or memory that may apply in clinical settings. In the case of depressive delusions, the unavailability of alternative hypothesis is driven by the process of negatively biased learning we described earlier: people have long ignored positive information about themselves or re-interpreted it with a negative spin, and this has not been integrated in their self-schemata.
Third, epistemic innocence applies to the adoption, not the prolonged maintenance, of delusional beliefs. The benefit consists in avoiding a problem that presents itself when the unusual experience or the new evidence is not yet made sense of, and can generate anxiety, stress, tension. Delusions appear as a response to a critically high point (a ‘tipping point’) of cognitive imbalance (or, quoting Jaspers, a ‘limit situation’), and their function is to reinstate the balance that has been lost.
From an epistemic standpoint, by reducing dissonance and preventing the disintegration of the self, depressive delusions help restore the person’s narrative identity. But the benefit is temporary. When the delusion becomes entrenched, as we showed both in the case of schizophrenic and depressive delusions, it becomes a new source of stress, anxiety, and tension.
According to the cognitive model, preserving consistency is a priority for human cognition (Festinger, 1957). People have a general tendency to restore coherence between their views, and between their views and the views of others, when this is compromised. If a person finds herself in a state of extreme cognitive dissonance, she needs to validate her prior beliefs and offset the cognitive dissonance emerging from those beliefs and newly acquired information. Mood-congruent delusions in depression play this key role. In particular, they may offer validation for one’s own intensively experienced guilt, shame, hopelessness, and dismay.
The claim that one is being spied upon in one’s own home – for example – matches the belief that one is not trustworthy and should be monitored at all times. Undoubtedly such delusions come with psychological costs. The delusional content is weaved with unpleasant, sometimes terrifying events, such as being watched, followed, or threatened. But, by reducing dissonance, delusions also offer psychological relief from dissonance–related anxiety. The delusion–induced anxiety that replaces the dissonance-related anxiety has the advantage of reinstating consistency: the negative beliefs about the self are confirmed.
Our hypothesis is compatible with the prediction-error theory of delusion formation (e.g. Fletcher and Frith, 2009; Corlett et al., 2007) which has already been applied to delusions in schizophrenia. According to this model delusions are formed in response to aberrant prediction-error signals, those signals that indicate a mismatch between expectation and actual experience (Miyazono et al., 2014). A prediction error happens when new incoming information does not match the person’s existing representations (schemata) and, therefore, cannot be successfully integrated in the person’s model of the world (assimilated). It indicates that the internal model of the world from which the prediction is derived is incorrect and needs to be updated. By updating the model in such a way as to minimize prediction errors, the person gains a better understanding of the world.
In the case of depressive delusions, the process differs as people err on the side of conservatism as opposed to revisionism. Instead of changing their representation of the world to match the unusual experience, validating new experience at the expense of previously acquired beliefs, people with depressive delusions preserve their representation of themselves despite the new information that conflicts with it, validating the existing self-schemata at the expense of the new information that is reinterpreted and distorted to fit the schema. When this happens, an adaptation process has to occur for the system to regain cognitive balance.
Here we suggest that mood-congruent, motivated delusions in schizophrenia are epistemically innocent. Here are the two conditions for the epistemic innocence of delusions: 1. Epistemic benefit: The adoption of the delusional hypothesis confers a significant epistemic benefit to a given agent at a given time, 2. No alternatives: Alternative hypotheses that would confer the same benefit are not available to that agent at that time.
Some qualifications are in order. First, what counts as an epistemic benefit may vary. One might say that adopting a delusion is epistemically beneficial if it contributes to the acquisition or retention of true beliefs, if it promotes the agent’s intellectual virtues, or if it is something an agent should be praised and not blamed for. One’s commitments in epistemology will affect the way in which the epistemic benefits are identified and described. In the case of depressive delusions, the main benefit seems to be the preservation of a coherent self-representation.
Second, different notions and degrees of unavailability can explain the failure to adopt a less epistemically costly hypothesis. This spectrum of possibilities reflects the nature of the limitations that the agent experiences in the relevant context, ranging from standard reasoning limitations affecting all human agents to deficits of perception, inference, or memory that may apply in clinical settings. In the case of depressive delusions, the unavailability of alternative hypothesis is driven by the process of negatively biased learning we described earlier: people have long ignored positive information about themselves or re-interpreted it with a negative spin, and this has not been integrated in their self-schemata.
Third, epistemic innocence applies to the adoption, not the prolonged maintenance, of delusional beliefs. The benefit consists in avoiding a problem that presents itself when the unusual experience or the new evidence is not yet made sense of, and can generate anxiety, stress, tension. Delusions appear as a response to a critically high point (a ‘tipping point’) of cognitive imbalance (or, quoting Jaspers, a ‘limit situation’), and their function is to reinstate the balance that has been lost.
From an epistemic standpoint, by reducing dissonance and preventing the disintegration of the self, depressive delusions help restore the person’s narrative identity. But the benefit is temporary. When the delusion becomes entrenched, as we showed both in the case of schizophrenic and depressive delusions, it becomes a new source of stress, anxiety, and tension.