On the dissociation between clustering and switching in verbal fluency: comment on Troyer, Moscovitch, Winocur, Alexander and Stuss.

Troyer, Moscovitch, Winocur, Alexander and Stuss (Neuropsychologia 36 (1998) 499) used a procedure originally introduced by Troyer, Moscovitch and Winocur (Neuropsychology 11 (1997) 138) for distinguishing two different components of verbal fluency-clustering and switching-in frontal and temporal-lobe patients. Application of this procedure yielded results that suggested a frontal-lobe switching deficit, but intact 'clustering'. I demonstrate here that the proposed procedure may lead to incorrect conclusions because it does not allow an unambiguous dissociation between a general reduction in processing speed and a selective switching deficit. Some implications of this critique for inferences about the neurocognitive components involved in verbal fluency tasks are discussed.