Most aspects of standard definitions have exceptions on which clinicians in practice agree. The notion that delusions like other psychotic phenomena are best viewed as lying on a continuum with normal beliefs is appealing but flawed (David, 2010). Delusions are multiply determined with many typical features but none essential and / or invariable. Delusions are more than the sum of their epistemic features. They are clinical phenomena that require a huge amount of contextual information before they can be understood (or before it can be concluded that they are un-understandable!). Sometimes a statement that would only be weakly eligible to be called a delusion, passes this threshold because the person expressing the utterance shows evidence of psychosis on other grounds. This 'reverse reasoning' would be wholly inadmissible epistemologically but is perfectly acceptable and indeed sensible in a clinical context.
The delusion should be considered to be akin to ‘a syndrome’ and has similar utility.
David AS (1999) On the impossibility of defining delusions. Philosophy, Psychiatry, & Psychology, 6 (1), 17-20.
David AS (2010) Why we need more debate on whether psychotic symptoms lie on a continuum with normality. Psychological Medicine, 40: 1935-42.