Clustering and Switching in Verbal Fluency in Parkinson's Disease

Thirteen patients with non-dementing Parkinson's disease (PD) were compared with 11 healthy controls on a semantic and a phonemic verbal fluency task. Analysis of the data examined total words produced as well as two additional components of verbal fluency: Clustering (generating words which share a semantic or phonemic similarity), and switching (the ability to shift between clusters). Overall, PD patients generated fewer words than the controls, and made less use of switching both in the phonemic and semantic fluency tasks. They did not differ from the controls in the amount of clustering. The results are consistent with the notion that patient groups in which the frontal lobes are compromised will perform poorly on switching tasks. Verbal fluency tasks are a standard part of the neuropsychological assessment of many neurological and neuropsychiatric disorders. Typically the participant is asked to name as many words beginning with a particular letter of the alphabet (phonemic), or to name as many examples of a specific category such as vegetables (semantic), in a fixed time period (Lezak, 1995). However, while it is generally accepted that fluency may suffer as a consequence of a range of different neuropathological conditions, it remains uncertain exactly which anatomical structures, and which cognitive processes, might underlie such deficits. One approach to clarifying these issues has been to study fluency deficits across different neurological conditions. For example, researchers have examined verbal fluency in patients with Alzheimer's disease (Hart, Smith, & Swash, 1988), Parkinson's disease (Hanley, Dewick, Davies, Playfer, & Turnbull, 1990), mild traumatic brain injury (Raskin & Rearick, 1996), and depression and schizophrenia (Crowe, 1992). This comparative approach has been useful in helping to reveal the differences in fluency that can arise with different diseases and their implications for our understanding of the brain structures involved. For example, researchers typically report a deficit on tasks of semantic (category) fluency amongst patients with Alzheimer's disease where the temporal lobes (and semantic memory stores) are likely to be compromised (e.g., Randolph, Braun, Goldberg, & Chase, 1993; Mickanin, Grossman, Onishi, Auriacombe, & Clark, 1994). By contrast, researchers studying Parkinson's disease (PD), where cortical impairment most likely involves the frontal lobes with consequent compromise of retrieval processes, typically report impairment on tasks of phonemic fluency but less so on tasks of semantic fluency (e.g., Bayles, Trosset, Tomoeda, Montgomery, & Wilson, 1993). Research on verbal fluency has also concentrated on the specific cognitive processes that might underlie verbal fluency, and that may be impaired when there is brain pathology. The most popular strategy has been where researchers have compared their participants' performances on phonemic and semantic fluency tasks. The assumption here is that the two tasks are qualitatively different and that performance differences may give clues about cognitive processes involved. Probably the most frequently reported finding here is that participants can produce fewer words on the semantic version of the task than the phonemic. For example, Troyer, Moscovitch, & Winocur, (1996,1997) reported this with younger and older healthy adults, and Raskin, Sliwinski and Borod (1992) with patients with PD and also their control participants. Auriacombe, Grossman, Carvell, Gollomp, Stern, and Hurtig (1993) have suggested four specific cognitive processes that could be involved in verbal fluency tests. These are: (a) attention and vigilance, (b) a lexical or semantic store, (c) a retrieval mechanism, and (d) a working memory that monitors items already produced. However, while these four putative component processes fit well with contemporary information processing accounts of human cognition, most neuropsychological studies on verbal fluency have only measured the final output, or total number of words produced. …