There are a few prominent eponymous syndromes in psychiatry; two of the most well-known are the Capgras delusion and the Fregoli delusion. These are delusional misidentification syndromes: In the Capgras delusion, the sufferer believes that a close relative has been replaced by someone, or is not who they say they are, despite them physically resembling the person they have replaced. The original description (Capgras & Reboul-Lachaux Reference Capgras and Reboul-Lachaux1923/1994) described the delusion as an “agnosia of individual identification” (p. 1) and not necessarily a symptom of false recognition, highlighting the separation between recognition and identification. In the Fregoli syndrome, the patient holds the delusional belief that one person is constantly changing his or her appearance and occupying different forms, thus appearing as different people. The Fregoli delusion was originally described in a 1927 publication as a complimentary antonym to Capgras syndrome (Courbon & Fail Reference Courbon and Fail1927/1994). In the Capgras delusion, the patient can recognize the similarity of the “imposter” to the close friend or relative and can recollect the facial detail of said person (Josephs Reference Josephs2007). An additional element for those suffering from Fregoli syndrome is that patients appreciate that the person looks different, but believe this is the same person despite the different superficial appearance (Langdon et al. Reference Langdon, Connaughton and Coltheart2014). Thus, both these syndromes reflect a dichotomy between recollection and familiarity. What does the new integrative memory model presented in the current target article by Bastin et al. suggest may be the issue in these psychiatric syndromes? Ostensibly, in the Capgras delusion there is intact recollection and a deficit in or absence of familiarity; while in the Fregoli delusion, there is a deficit in recollection and over-attribution of familiarity. The deficits could occur at the level of the initial or core processing, or at the higher-order attributional stage where recollection and familiarity are combined and contextualized.

These delusional misidentification syndromes commonly occur in schizophrenia and psychosis (Förstl et al. Reference Förstl, Almeida, Owen, Burns and Howard1991) affecting around 15% of those suffering from schizophrenia (Feinberg & Roane Reference Feinberg and Roane2005; Salvatore et al. Reference Salvatore, Bhuvaneswar, Tohen, Khalsa, Maggini and Baldessarini2014). However, delusional misidentification syndrome has also been attributed as a symptom of many other disease states, including dementia, epilepsy, Parkinson's disease, trauma, and other organic brain diseases (Oyebode & Sargeant Reference Oyebode and Sargeant1996; Pandis & Poole Reference Pandis and Poole2017). Neuroimaging has highlighted the role of organic brain dysfunction in delusional misidentification syndrome (Atta et al. Reference Atta, Forlenza, Gujski, Hashmi and Isaac2006). These studies of delusional misidentification syndrome have highlighted prominent dysfunction in frontal cortical regions, with a focus on right-hemispheric change, and in temporoparietal cortical regions, with an emphasis on left-hemisphere change. A few studies have suggested parahippocampal atrophy. Specific studies have demonstrated Fregoli's delusion associated with right-frontal and left-tempo-parietal contusions following trauma (Feinberg et al. Reference Feinberg, Eaton, Roane and Giacino1999). Meta-analyses of delusional misidentification syndrome point to the involvement of the right frontal lobes (Atta et al. Reference Atta, Forlenza, Gujski, Hashmi and Isaac2006; Feinberg & Roane Reference Feinberg and Roane2005), with other changes observed in the left temporal lobes (Edelstyn & Oyebode Reference Edelstyn and Oyebode1999; Feinberg et al. Reference Feinberg, Eaton, Roane and Giacino1999; Huang et al. Reference Huang, Liu and Yang1999; Signer Reference Signer1994). Further, neuroimaging data in delusional misidentification syndrome and schizophrenia have demonstrated structural volume reduction in the frontotemporal area of the brain (Turkiewicz et al. Reference Turkiewicz, Zanetti, Zung and Cordeiro2009), as well as reductions in structural magnetic resonance in the right frontal lobe (Coltheart et al. Reference Coltheart, Langdon and McKay2007). A few studies showed damage to the right fusiform gyrus and para/hippocampal atrophy indicating temporal lobe deficiencies (Hudson Reference Hudson2000).

In the integrative memory model, the fluency heuristic establishes familiarity, while the attribution system underpins recollection. These combine with the core systems generating the memory trace that is made available to these familiarity and recollection systems to ensure accurate memory-related decisions. The imaging data in delusional misidentification syndrome largely implicate dysfunction at the level of the fluency heuristic and attribution, based on the prefrontal cortex and possibly its connections with perirhinal regions, rather than reflecting any core dysfunction in the entity or relational processes associated more with the hippocampus and perirhinal regions. The model appears to lack parsimony as it proposes that a dissociation between familiarity and recollection can occur through damage to the hippocampus or the perirhinal cortex – but it is not clear how it is possible to distinguish the consequences of this from any prefrontal cortical dysfunction which will impact attribution and lead to misrecognition and impaired familiarity.

In summary, the structural and functional anomalies found in patients suffering from Capgras and Fregoli misidentification syndromes, that demonstrate aberrant integration of recollection and familiarity, fit with a dysfunction at the level of the attributional system rather than a core representational deficit in Bastin et al.’s new model. However, the authors propose that similar dissociations are also possible with a core dysfunction in hippocampus or perirhinal cortex, which suggests that the model may lack parsimony and potentially fits less well with the available data on dissociation.