Modality-specific effects in inhibitory mechanisms: The interaction of peripheral and central components in working memory Paola Palladino, a Nicola Mammarella, b and Tomaso Vecchi a a Dipartimento di Psicologia, Universit  a di Pavia, Italy b Dipartimento di Psicologia Generale, Universit  a di Padova, Italy Accepted 7 May 2003 Abstract Working memory is a complex system assumed to encompass both storage and processing components. At the same time, several authors have suggested the existence of separate peripheral modality-specific subsystems, in particular a verbal and a visuo-spatial mechanism. In the present research, we explored the hypothesis that modality-specific mechanisms may also operate in the central processing component. Inhibitory control has traditionally been considered as a typical function of the central executive, devoted to selecting, processing, and eliminating irrelevant information. We designed verbal and visuo-spatial tasks involving similar proce- dures and comparable levels of inhibitory control. Results indicate that accuracy profiles are similar in the two conditions; however, the patterns of data measuring inhibitory control, i.e., percentage of intrusion errors, are opposite in the case of verbal (intrusions decreased in more complex trials) and visuo-spatial (intrusions increased in more complex trials) material. These data are consistent with theoretical models of working memory suggesting that modality-specific effects emerge even in central processes and that both material-dependent and process-dependent features must be considered in working memory tasks. Ó 2003 Elsevier Inc. All rights reserved. 1. Introduction It has been repeatedly demonstrated that working memory (WM) processes underlie complex cognitive tasks, such as reasoning, problem solving, and reading comprehension (Daneman & Merickle, 1996; De Beni, Palladino, Pazzaglia, & Cornoldi, 1998; Passolunghi, Cornoldi, & De Liberto, 1999). Although it is well es- tablished that in WM tasks participants are required to store and contemporarily process information (Badde- ley, 1986), the nature and specific characteristics of the memory components involved in the different tasks re- main highly unspecified. Hasher and Zacks (1988) suggested that efficient performance in WM tasks implies control over the ac- tivation of processed information. In fact, WM tasks require selective recall of target information only. Of all the available information, only target information must be maintained and recalled. Therefore, a significant amount of information previously processed should be disregarded (inhibited). Poor inhibitory control of ir- relevant information could reduce WM efficiency: irrel- evant information employs memory resources thus reducing the overall capacity of the system. Several studies have found a significant relationship between the recall of irrelevant information (intrusion errors) and WM performance in verbal tasks. Intrusion errors appear to be more frequent in participants with WM problems, sometimes associated with difficulties in complex cognitive tasks (De Beni et al., 1998; Palladino & De Beni, 1999; Passolunghi et al., 1999). Intrusion errors may be interpreted as due to a difficulty in re- ducing activation of irrelevant information (inhibitory control difficulty). Low span participants, besides pro- ducing a higher number of intrusion errors, also showed an advantage in re-learning previously irrelevant infor- mation (Rosen & Engle, 1998). Palladino, Cornoldi, De Beni, and Pazzaglia (2001) showed that conditions re- quiring greater updating (inhibition of previously pre- sented information) induced a higher production of intrusion errors in a verbal WM task. Several studies have investigated verbal WM, but little attention has been paid to examining inhibitory control in visuo-spatial working memory processes. The Brain and Cognition 53 (2003) 263–267 www.elsevier.com/locate/b&c 0278-2626/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/S0278-2626(03)00123-4