Proper name anomia and proper name sparing are unusual disorders characterized by a selective inability, or sparing, to retrieve proper names as opposed to common names. They may eventually demonstrate that these two categories of names are independently processed in the brain at the semantic or lexical/orthographic level. Yet, differences in patient assessment make comparison between cases difficult to interpret and raise alternative explanations to those dissociations, such as that they may be accounted for by the level of difficulty of test stimuli.
We describe two individuals, of identical age, gender and cultural background, with a complementary pattern of impairment of name retrieval following a left hemisphere stroke. ACB had an aphasic disorder with impaired naming, sparing proper names, while JFJ had normal language abilities and semantic knowledge about people, but a marked anomia for people's names. Patients were studied using the same material, thus overcoming methodological constraints pointed to previous work. These cases provide evidence of a double dissociation between proper and common names at the lexical access level. In addition, JFJ had a normal ability to retrieve historical and religious names suggesting that either these names are less difficult to retrieve or that they may segregate within the proper name lexicon.
Lesion of the left temporal pole was also dissociated. It was damaged in ACB but spared in JFJ, contradicting its crucial role in proper name processing.
The functional architecture of the cognitive system has been largely inferred from the study of individuals with acquired brain lesions, in particular in what concerns the neural organization of the lexical-semantic knowledge (Caramazza, 1998 and Hart et al., 1985; Warrington & Shallice, 1987).
Regarding the neural representation of proper names, several case studies have shown that, following a cerebral lesion, people's names can be selectively affected (Hittmair-Delazer, Denes, Semenza, & Mantovan, 1994; Luchelli & De Renzi, 1992; Semenza and Zettin, 1988 and Semenza and Zettin, 1989) or spared (Cipolotti, McNeill, & Warrington, 1993; Lyons, Hanley, & Kay, 2002; Schmidt and Buchnan, 2004; Semenza & Sgaramella, 1993) compared to common names, an evidence suggesting that they are processed by distinct neuronal networks. These dissociations have been described at different cognitive loci (semantic level, phonological access, orthographic output, etc.), but mostly at the semantic comprehension level (Brédart, Brennen, & Valentine, 1997). However, the basis of such dissociation is still uncertain, due to the different nature of the experimental stimuli used and the lack of direct comparisons between cases.
In addition to individual case reports, data supporting a segregation between proper and common names has been provided by other sources, such as the patterns of lesion observed in series of patients (Damasio, Grabowski, Tranel, Hichwa, & Damasio, 1996), the record of daily errors produced by healthy individuals (Young, Hay, & Ellis, 1985), ERP studies (Proverbio, Lilli, Semenza, & Zani, 2001) and brain activation studies (Damasio et al., 1996, Gorno Tempini et al., 1998, Grabowski et al., 2001 and Rotschtein et al., 2005). The latter have also suggested that the left temporal pole may play a crucial role in proper name retrieval, although this has been questioned by systematic reviews of published cases (Semenza, Mondini, & Zettin, 1995; Yasuda, Nakamura, & Beckam, 2000).
The dissociation between proper and common names has been generally explained by their semantic uniqueness. While common names refer to concepts, proper names refer to single entities, do not contain attributes by themselves and are arbitrary, being considered “pure referential expressions,” i.e., labels which function is merely to refer to entities to which they are associated (Kripke, 1980, Semenza, 2006, Semenza, 1997 and Semenza and Zettin, 1989). In contrast, common names refer to a set of attributes that are shared by multiple entities within the same concept. As a consequence of that, proper names would have rather fragile associations to their unique reference, while common names would be represented in widespread neuronal networks that could be accessed and activated from different cognitive entries. Still another difference between these two categories of names concerns the alternatives offered to their retrieval (Brédart, 1993). Common names can often be substituted by synonyms while proper names become unspecific or unrecognizable, if partly substituted or omitted. All these features make proper names particularly difficult to learn and to retrieve by healthy subjects (Cohen, 1990 and McWeeny et al., 1987), especially during normal ageing (Evrard, 2002 and Pluchon et al., 2002) and in the early stages of dementia (Semenza, Mondini, Borgo, Pasini, & Sgaramella, 2003).
We report two patients with naming impairments following a single left hemisphere stroke, who presented a contrasting performance between proper and common names, apparently at the access stage to the phonological lexicon. This double dissociation cannot be accounted for by the difficulty or complexity of the stimuli, since those were the same for both patients, nor by cultural differences between them, for patients were matched for age, gender and education. To our knowledge, this is the first report of a complementary impairment of proper names at this cognitive level. Lesion localization was also dissociated in what concerns the involvement of the temporal pole, and contrary to current theories on its role in proper name retrieval (Grabowski et al., 2001).
