Tuesday 26 January 2021

Against Defining Delusion

Today's post is by Sam Wilkinson. You can read Sam's recently published paper, Expressivism about Delusion Attribution, in the European Journal of Analytic Philosophy. It appeared in a special issue on the Bounds of Rationality.

Sam Wilkinson

What is delusion? While we can point to paradigmatic cases, we have struggled to produce an uncontentious definition of delusion. In my paper, I argue that we shouldn’t have been trying to define delusion in the first place, and that it becomes clear why, once we reflect on the sort of concept that delusion is.

Delusion attribution, e.g. saying “This person is delusional”, is not (fully) fact-stating. It is not like saying “This person is 6ft tall”. It is fundamentally an evaluation. Some evaluations involve failing to adhere to an objective benchmark, while others are more fundamentally, irreducibly evaluative. One way of thinking about these fundamental kinds of evaluations is as expressive, rather than descriptive.


To simplify somewhat, delusion attributions are more akin to expressions of folk-epistemic bafflement. Why think that this is so? And what are the consequences of this?

First of all, it seems like delusion expresses a negative evaluation. Consider a hyperbolic use among friends (“You’re delusional if you think Manchester Utd. will win the Premiership!”). Now, one might object that, while this is an expressive use of “delusion”, this is not at all how psychiatrists use the term. Nevertheless I would insist that the calm, institutional use of the term only masks its true nature.

Secondly, as with expressivists in meta-ethics, who are motivated by a metaphysical unease with postulating moral properties we might be similarly reluctant to countenance properties of “delusionality”.

What are the consequences of the view?

First, an inability to define is to be expected. Why think that all of the many things that arouse negative reactions from our folk-epistemic sensibilities can be regimented under a strict definition?

In a related manner, if you try to characterize the sorts of things that get called delusions, you get a sort of disjunctive norm pluralism. Put simply, there are different kinds of folk-epistemic badness, but once a threshold is reached, you call that thing a delusion. Contrast, for example, reverse Othello syndrome with a bizarre delusion in the context of psychosis. You might, for example, think about the implausibility of the content. Here, the bizarre delusion (“I am the left foot of God”) scores highly, so highly in fact that you don’t even need to ask yourself about the individual’s grounds for the claim. (Though interesting, it wouldn’t remove delusional status.) In contrast, if somebody said “My wife is being faithful to me”, this is true (one hopes) of many people, and it could only be an examination of their personal situation, their resistance to evidence, etc. that would then tip them into delusional territory.


Since delusion attribution is an expression of bafflement, understandability takes precedence over rationality. A mother who is reluctant to believe that her son is a murderer may exhibit profound levels of epistemic irrationality, but we understand that people are reluctant to admit the guilt of loved ones. Indeed, we’d find an impartial acceptance of the evidence more troubling!

What does this mean for delusions researchers? Not much. Keep up the great work on paradigm cases of delusion. Just don’t worry about defining it.

Tuesday 19 January 2021

Disorders of Agency on a Spectrum

Today's post is by Valentina Petrolini (University of the Basque Country – UPV/EHU). Here she talks about a recent paper she wrote, “Too Much or Too Little? Disorders of Agency on a Spectrum” published open access in a special issue of the European Journal of Analytic Philosophy on Bounds of Rationality.

Valentina Petrolini


“Rock You like a Hurricane” has been playing on repeat in my head since yesterday. I am unsure where it came from, although I am afraid a binge session of Stranger Things might have something to do with it. Despite my attempts, getting rid of the song proves surprisingly difficult. In my paper I characterize these episodes as mild cases of hypoagency. An action – in this case a mental one – is attributed to an agent, who is unsure about having initiated it and lacks a robust sense of control over it.

Some instances of hypoagency – such as having an 80’s song stuck in your head – strike us as relatively innocent. We may even imagine circumstances in which similar mind-wandering episodes may be beneficial. Assume that I have an important interview coming up and I cannot concentrate because my thoughts keep drifting away. Although my fantasizing might take a pessimistic turn, it might also allow me to come up with original ideas that help me succeed in the interview.

In an interesting study, Baird and colleagues show that some forms of mind-wandering might indeed support problem-solving. Other experiences of hypoagency are undoubtedly less benign. Think about phenomena such as auditory verbal hallucinations (AVH), in which a person loses grip over her own thoughts and experiences them as alien. Longden describes her voice-hearing experience along these lines: the first voices sound like “a running commentary”, but over time they grow in number and intensity and start issuing threats and commands over which she has no control.

However, there are also situations in which people experience hyperagency. A clinically-relevant example concerns cases of pathological guilt, commonly experienced by people diagnosed with schizophrenia. For instance, Saks talks about being filled with anxiety when reading the newspaper, because she would blame herself for every violent crime reported in the area. Some cases of hyperagency do not qualify as pathological. 

An interesting phenomenon in this sense is false confessions, in which people take responsibility for crimes that they have not committed. Although the idea of innocent people willing to face legal charges appears counterintuitive, studies in forensic psychology show that false confessions are relatively frequent. A famous case is portrayed in DuVernay’s series When They See Us and involves the men who came to be known as the “Exonerated Five”.

How should we think about different cases of disrupted agency? One important distinction concerns self-attribution versus feeling of agency. While self-attribution may be correct or incorrect, the feeling of agency comes in degrees. Extreme cases of hypoagency – such as AVH – exemplify situations in which self-attribution is incorrect and the subject lacks a robust feeling of agency. By contrast, in mind-wandering the perceived lack of agency usually fails to be accompanied by misattribution: even if I can’t stop thinking about the Scorpions song, I do not perceive it as externally generated. 

Things are more complex with hyperagency. Both pathological guilt and false confessions exemplify situations in which self-attribution is incorrect and the subject reports a strong feeling of agency. The difference between these cases is thus likely to be one of degree, where quantitative factors such as duration or intensity may be taken as reliable indicators.

Tuesday 12 January 2021

Reflections about electroconvulsive therapy

Today's post is by Emiliano Loria (Universit√† La Sapienza, Roma). Here he summarises a recent paper he wrote, "A desirable convulsive threshold: Some reflections about electroconvulsive therapy", published open access in a special issue of the European Journal of Analytic Philosophy on Bounds of Rationality.

Emiliano Loria

Long-standing psychiatric practice confirms the pervasive use of pharmacological therapies for treating severe mental disorders. Nevertheless, we are far from triumphal therapeutic success. Despite the advances made by neuropsychiatry, this medical discipline remains lacking in terms of diagnostic and prognostic capabilities when compared to other branches of medicine. 

An ethical principle remains as the guidance of therapeutic interventions: improving the quality of life for patients. Unfortunately, psychotropic drugs and psychotherapies do not always result in an efficient remission of symptoms. I corroborate the idea that therapists should provide drug resistant patients with every effective and available treatment, even if some of such interventions could be invasive, like Electroconvulsive Therapy (ECT).

ECT has an almost centennial history that began in Rome (Italy) in 1938, at the Clinic of Nervous and Mental Diseases, run at the time by (psychiatrist) Ugo Cerletti. ECT still represents one of the most important and controversial therapeutic discoveries in the field of psychiatry. ECT carries upon its shoulders a long and dramatic history that should be better investigated to provide new insights. 

From the examination of the Archives of Pediatric Neuropsychiatry in Rome - a section of the Roman Clinic specifically dedicated to minors - I discovered t the first child ever administered with ECT (September 18, 1940), a 7-year-old boy diagnosed with “dementia praecocissima”, a diagnostic category introduced by Sante De Sanctis, who was Ugo Cerletti’s predecessor to the direction of the Roman Clinic, as well as the one who established the first department of Neuropsychiatry.

ECT has attracted renewed interest in recent years. This is due to the fact that antidepressant drugs in younger patients show often scarce effectiveness and unpleasant side-effects. Moreover, thanks to modern advances, ECT may work as a successful form of treatment for specific and rare cases, such as severe depression (with suicide attempts) and catatonia. 

When pharmacotherapy fails to improve depressive symptoms, then, response rates of about 50–60% can be achieved by ECT. For this reason, particularly in depressed patients at high risk of suicide, ECT should be recommended earlier than its conventional “last resort” position. In fact, the risks of suicide have been shown to relieve quickly through ECT, when administered in continuity with previous treatments, that are essential to sustain its benefits.

The ethical puzzle that I raise is the following. Is it possible to administer a therapy to help severely suffering patients (be them adults or minors)? If the answer is yes, at least for some types of severe diseases, why should we prolong the severity of the symptoms by making the life of patients and their family members unbearable? One therapy for the improvement of some severe psychotic symptoms exists and is practicable. ECT is such a therapy. In this sense, ECT does not constitute an alternative model of treatment, but an additional therapeutic tool that does not replace, but rather integrates pharmacotherapy and psychotherapy.

Tuesday 5 January 2021

Delusions in the two-factor theory: pathological or adaptive?

Today's post is by Eugenia Lancellotta (University of Birmingham). Here she talks about a recent paper she wrote, "Delusions in the two-factor theory: pathological or adaptive?", published open access in a special issue of the European Journal of Analytic Philosophy on Bounds of Rationality.

Eugenia Lancellotta

Are delusions pathological, adaptive, or both? I investigated this issue with Lisa Bortolotti. We framed the question in the context of one of the most popular theories of delusion formation and maintenance: the two-factor theory.

Two-factor theories hold that the formation and maintenance of delusions involve two factors. Factor 1 is usually a neuropsychological impairment, while Factor 2 is a cognitive deficit or bias. While two-factor theorists agree on the broad two-factor architecture involved in the formation and maintenance of delusions, they disagree on some aspects of it. Coltheart and McKay are among the most prominent two factor theorists. 

While for Coltheart, Menzies and Sutton (2010) Factor 2 is a cognitive deficit that only gets activated in the maintenance stage of delusions, for McKay (2012) it is a bias that is already present when delusions are adopted. In other words, for Coltheart delusions are initially a rational explanation of the anomalous feelings or experiences engendered by Factor 1, but they then become problematic when maintained in the face of counterevidence. For McKay instead, adopting the delusion in the first place is already problematic. Therefore, according to McKay, a bias in reasoning – though not a deficit - must already be present in the adoption stage of delusions.

How do these theories relate to the notion of pathology? We argue that in the Coltheart model, delusions are pathological both on a normativist and naturalist view. On a normativist view, delusions are pathological because they disrupt psychological functioning, while on a naturalist view, delusions are pathological because they are due to a cognitive dysfunction (Factor 2 in the Coltheart model). For McKay, delusions are not pathological on a naturalist view, because their formation and maintenance are due to a bias rather than to a dysfunction, while they can be pathological on a normativist view, because they can disrupt psychological functioning.

What about the relationship that such theories entertain with adaptiveness? Delusions are adaptive if they are designed to act as an emergency mechanism, providing psychological or biological benefits in the face of adversities, traumas or cognitive impairments. On Coltheart’s reading, maintaining delusions in the face of counterevidence could be a response to a crisis that prevents the cognitive system from collapsing, so delusions might be adaptive. On McKay’s model, adopting some delusions could be a response to a crisis that prevents the cognitive system from collapsing, so delusions could be adaptive when adopted. This is compatible with those delusions being the outcome of a cognitive bias.

The conclusion of our investigation is somewhat surprising. In the McKay model, some delusions can be pathological and adaptive, though not at the same time. Adaptive, because they prevent the person’s cognitive system from breaking down when adopted; pathological, on a normativist account, because they disrupt the person’s psychological functioning in the long-term. However, in the Coltheart model, delusions cannot be adaptive and pathological, because the fact that they are the outcome of a dysfunction (pathological in a naturalist sense) is incompatible with the possibility that they are the outcome of an emergency mechanism which breaks by design.