The case of a patient with selective left hemispheric medial and lateral temporal lobe damage is described. The patient was of slightly supra-average intelligence and had no problems in normal memory functions, but was severely anomic with respect to people's names. One month post-onset, this deficit held for names of colleagues and friends she had gotten to know during the last 10 years prior to the infarct and for all names confronted with post-infarct. On the other hand, learning of face–name associations was preserved and was independent of the ability to generate context-specific information for the subjects whose names were requested. The results support the existence of category-specific naming impairments, and, moreover, indicate a deficit that has to be differentiated with regard to memory systems. A time-limited, but prolonged engagement of interconnected left medial and adjacent lateral temporal lobe structures in ecphory is stressed for context-restricted information such as proper names.
Evidence from a number of sources including positron-emission-tomographic studies 18, 26, 35, 36, 42 and 58and single case reports 12, 13, 14 and 50makes it likely that circumscribed cortical regions represent specific semantic attributes. After certain kinds of brain damage patients may no longer be able to name and describe living objects such as a snake or a crocodile, while they are still able to name and describe objects like a tent or a compass 11 and 61. A special class of semantic termini is defined by proper names and—as a specific subset—by people's names, and may be selectively impaired by cortical damage 5, 25, 30, 37, 46, 47, 48 and 49.
Possibly dependent on specific features of the impairment in name reproduction, left thalamic 6, 30 and 39, left temporal 22, 37 and 49, left parieto-occipital [47]and left-hemispheric fronto-temporal damage 25 and 48have been found to result in impairment for proper names. While lesion data are divergent, Damasio and colleagues [9]conducted word-retrieval experiments with both brain damaged patients and normal subjects undergoing positron-emission-tomography (PET). For both groups of subjects, retrieval of words signifying concrete entities (names for persons, animals, tools) depended on different regions of the temporal lobe, whereas naming actions activated the left frontal cortex. In a study likewise based on PET data, Martin et al. [36]found that naming pictures of animals was associated with bilateral activations of the ventral temporal lobe, Broca's area and the medial occipital lobe.
With regard to the nature of the deficit, anomia may or may not be embedded in the context of more widespread amnesia. Based on different assumptions, numerous authors have suggested that anomia for proper names constitutes a retrieval deficit 15 and 22, or, more specifically, an inability to form—for retrieval—new associations between the meanings of words and their phonological forms [49].
Several further findings have helped to clarify the cognitive aspects involved in naming people. Flude and colleagues [17]assumed that names are stored separately from semantic information about people, and Damasio et al. [9]even hypothesized segregated neural systems subserving conceptual and word-form knowledge for the same stimulus. From experimental studies with normal subjects, Valentine et al. [57]concluded that name recognition units mediating between word recognition and the access to identity-specific semantics exist. Similarly, Fery and colleagues [15]presumed that anomia is caused by an impaired access to intact phonological representations from semantic memory. On the other hand, as there is evidence that names (e.g. `Baker') with name–occupation homophones are more difficult to recall than occupations (e.g. baker) [38], this disadvantage for proper names cannot be attributed to differences in the phonological word-form [7], but has to be discussed in the context of word meaning. Indeed, proper names seem to be unique because of their low meaningfulness; they therefore work as `pure referring expressions' [48]without describing any properties or attributes of the corresponding persons [25]. Moreover, Burton and Bruce [4]emphasized the uniqueness of a name to a given person. A survey of current models and a discussion of recent cases can be found in Cohen and Burke [8].
We here report a new case of a selective anomia for people's names focusing on its neuronal correlates and on the nature of the distinct impairments of different memory systems.
