Editorial Automatic and controlled language processes: a special issue of the Journal of Neurolinguistics The distinction between automatic and controlled cognitive processes is a “classic” in contemporary psychology. Since the publication of Posner and Snyder (1975) and Shiffrin and Schneider (1977), several properties have been proposed to characterize these two types of mental operations. Automatic processes are fast-acting, they occur without intention or awareness, and they do not use limited-capacity resources. On the other hand, controlled processes are slower, they are under the strategic control of the subject, and they have a cost. This dichotomy has been challenged, and it seems clear that the two processes represent two extremes of a continuum (Cohen, Dunbar, & McClelland, 1990; Hampson, 1989). In the approach adopted by Shiffrin and Schneider (1977), automaticity is an acquired characteristic: mental operations that are sufficiently practiced are performed more quickly and accurately, they also undergo qualitative changes. Automaticity is commonly conceived as the result of practice. On the other hand, in a more recent approach, automaticity has become a central concept to consider the learning process itself. Here, we refer to the ability to learn without awareness: the implicit learning (cf. Cleeremans, Destrebecqz, & Boyer, 1998, for a review). Implicit learning is usually thought of as an automatic process. The two perspectives are considered in the papers collected in this special issue of the Journal of Neurolinguistics. The first two papers (Perruchet and Peereman; Dominey and Dodane) are mainly concerned by the automaticty of the first language learning process. They are followed by two papers (Franck, Cronel-Ohayon, Chillier, Frauenfelder, Hamann, Rizzi, and Zesiger; Rigalleau and Caplan) that consider developmental or acquired language disorders, both focusing on the hypothesis of a deficit in syntactic agreement processes. These processes are assumed to be automatic in normal adult speakers. Finally, two papers (Kotz and Elston-Gu ¨ ttler; Phillips, Segalowitz, O’Brien, and Yamasaki) examine the role of automaticity in second language (L2) acquisition, both using event-related potentials to examine how the variable speed of linguistic processes can distinguish more or less proficient L2 learners. Perruchet and Peereman compare two views about how the sensitivity to statistical regularities into the linguistic input can emerge. These authors note that the connectionnist approach (Elman, 1990) implies an ad hoc statistical device devoted to allowing 0911-6044/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0911-6044(03)00063-0 Journal of Neurolinguistics 17 (2004) 93–96 www.elsevier.com/locate/jneuroling