نام پریشی اسم خاص: یک مورد با دانش حفظ نخستین حرف

In this article we describe the case of GC, a woman affected by severe proper name anomia due to progressive brain atrophy that mainly affected the left temporal pole. Proper name comprehension and semantic knowledge about the people she was unable to name were normal. GC showed a sparing of initial letter knowledge of proper names, while other phonological characteristics were not equally available. At a later stage of her illness, the naming impairment began to affect common names as well as proper names, though at a lesser extent. Whereas there was no category effect between names of animate and inanimate stimuli, we observed a relative sparing of first letter knowledge selectively for animate categories, although less marked than with proper names. This case is discussed within the theoretical framework of two-stage models of name production. Knowledge of the initial letter of proper names supports the psychological reality of the ‘‘phonological address’’ as a preliminary stage of the production of this class of names. Moreover, the qualitative similarity between errors observed with proper names and with names of animate objects suggests that the production of names belonging to these classes may conform, at least in part, to analogous algorithms.

Patients affected by proper name anomia show a specific and consistent neuropsychological profile: they retain semantic information about people, but cannot retrieve the phonological form of their names. This category-specific disorder also participates in a double dissociation: selective preservation of proper names has been described by McKenna and Warrington [20], Semenza and Sgaramella [28]and restricted to the written modality by Cipolotti et al. [8]. This would rule out the hypothesis of an intrinsic difficulty associated with proper names. Hypotheses advanced to interpret proper name anomia have taken into account the special locus of proper names in the linguistic scenario 15 and 11. Proper names are considered ‘‘purely referring expressions’’ without implying any properties or attributes. A special vulnerability would therefore ensue from the ‘‘arbitrary link’’ between people and their names, despite the presence of the double dissociation noted above. Therole played by the initial segment of a name is relevant both from the clinical and the theoretical perspective. In a clinical survey, Goodglass et al. [14]found that, for a certain percentage of the names that aphasics could not produce, there was sparing of first letter knowledge; this occurred more frequently in conduction aphasics (about 30%) than in Wernicke’s and anomic aphasics. On the other hand, the ‘‘tip-of-the-tongue’’ phenomenon, observed also with normals, has been the subject of several inquiries [4]. ‘‘Tip of the tongue’’ affects proper names much more often than common names and first letter knowledge is often available to the subject during this transient word retrieval impairment. Stillfrom a clinical perspective, a classical facilitation technique in naming rehabilitation is to furnish the first phoneme of a name. Semenza and Sgaramella [28]described a patient, RI, affected by a severe deficit of language production and naming: he showed successful proper name retrieval only when given the first sound of the name as a cue. Semenza and Sgaramella suggested that phonemic cueing would be selectively more efficient (or even ‘‘sufficient’’) to foster proper name retrieval, in this case because the relationship between name and reference is very straightforward (‘‘one-to-one’’). Beyond the case of proper names, other theoretical approaches may also be relevant for a general interpretation of the role of the first segment in name retrieval. Both clinical studies and the errors observed in normals have provided evidence for current models of name production, as discussed by Howard [16]. We will concentrate here on the two-stage model of name production 18 and 5, in particular Butterworth’s model [5], as it gives a definite role to the first segment of a word. This model hypothesizes the existence of two separate loci where the information necessary for word production is stored: a ‘‘semantic lexicon’’ and a ‘‘phonological lexicon’’. Whereas the latter contains the explicit phonological form of the word, the former consists of semantic entries plus some fragments of the phonological structure of each word (the ‘‘phonological address’’) necessary for retrieving the appropriate representation in the phonological lexicon. The information contained in this ‘‘address’’ would include the first segment, the number of syllables, and the stress pattern of each word. According to Butterworth [5], the ‘‘address’’ can yield partial information about the phonological form of the word when the full phonological form is not accessible or is destroyed. Some case descriptions seem to add clinical evidence to the hypotheses reported above. For instance, the patient EST, described by Kay and Ellis [17]presented with relatively spared comprehension but a naming deficit due to a left hemisphere neoplasia. In many cases, this patient appeared to know little or nothing of the target name. A phonemic cue did not significantly enhance the retrieval of the full phonological form, but frequently facilitated the production of some phonological information. In other cases the patient could spontaneously retrieve partial phonological information but not the whole form. Unfortunately, the status of proper names was not extensively examined in this patient. In this article we describe the case of GC, a further patient affected by severe proper name anomia. The interest of this case is that, besides an almost complete sparing of semantic information, GC also showed preservation of knowledge of the initial letter of the names she was unable to retrieve: this acquired a special salience in her naming attempts. Her naming deficit was initially confined to proper names, but later involved also common names.

In April 1995 all the tests proved to be normal (Table 1). In particular, GC did not show language deficits other apart from proper name anomia. The neuropsychological examination was repeated on the following testing sessions, and no defects became apparent, with the lone exception of object naming to pictures from both living and non-living categories. Table 1. GC’s general assessment GC’s score Normal score Reference Intelligence and general ability MODA 92.2 ⩾89 [3] Raven PM47 matrices 36 ⩾18 [24] Memory Digit span 6 ⩾3.75 [23] Corsi span 4.25 ⩾3.50 [23] Paired associate 8.25 ⩾6.50 [21] Word lists 8 ⩾8 [21] Supra-spanlearning 6.64 ⩾5 [7] Language Token test 32 ⩾29 [10] Word comprehension 80/80 [6] Picture naming 29 ⩾28.25 [23] Naming to description 29.25 ⩾29 [23] Verbal fluency by letter 19 ⩾17 [23] Semantic fluency 25 ⩾25 [23] Perception Benton line orientation 28 ⩾17 [2] Benton face recognition 50 ⩾39 [2] Face perception Face familiarity −1.51 <1.42 [13] Famous faces −0.21 <1.32 [13]