www.elsevier.com/locate/brainres Available online at www.sciencedirect.com Research Report Emotion word recognition: Discrete information effects first, continuous later? Benny B. Briesemeister a,n , Lars Kuchinke b , Arthur M. Jacobs a,c a Freie Universität Berlin, Allgemeine und Neurokognitive Psychologie, Center for Applied Neuroscience, Habelschwerdter Allee 45, 14195 Berlin, Germany b Ruhr-Universität Bochum, Experimentelle Psychologie und Methodenlehre, Universitätsstraße 150, 44780 Bochum, Germany c Dahlem Institute for Neuroimaging of Emotion (D.I.N.E.), The Cluster of Excellence ‘Languages of Emotion’, Habelschwerdter Allee 45, 14195 Berlin, Germany article info Article history: Accepted 29 March 2014 Available online 5 April 2014 Keywords: Discrete emotions Affective dimensions Sequential affective processing EEG Lexical decision abstract Manipulations of either discrete emotions (e.g. happiness) or affective dimensions (e.g. positivity) have a long tradition in emotion research, but interactive effects have never been studied, based on the assumption that the two underlying theories are incompatible. Recent theorizing suggests, however, that the human brain relies on two affective processing systems, one working on the basis of discrete emotion categories, and the other working along affective dimensions. Presenting participants with an orthogonal manipulation of happiness and positivity in a lexical decision task, the present study meant to test the appropriateness of this assumption in emotion word recognition. Behavioral and electro- encephalographic data revealed independent effects for both variables, with happiness affecting the early visual N1 component, while positivity affected an N400-like component and the late positive complex. These results are interpreted as evidence for a sequential processing of affective information, with discrete emotions being the basis for later dimensional appraisal processes. & 2014 Elsevier B.V. All rights reserved. 1. Introduction Two main conceptions have been proposed to best describe human emotions, each being in accordance with convincing empirical data. On the one hand, a class of theories assumes that emotions are processed along a limited number of affective dimensions (Russell, 2003; Wundt, 1896). The ‘core affect’ theory (Barrett and Bliss-Moreau, 2009; Russell, 2003; 2005; 2009), for example, assumes that emotions are “grounded in continuous and fluctuating affective states described as pleasant or unplea- sant, with some level of arousal” within the core of the body (cf. Wilson-Mendenhall et al., 2013, p. 1). Within this class of theories, two affective dimensions, i.e. valence (ranging from a pleasant to an unpleasant pole) and arousal underlie human http://dx.doi.org/10.1016/j.brainres.2014.03.045 0006-8993/& 2014 Elsevier B.V. All rights reserved. n Correspondence to: Department of Psychology, Freie Universität Berlin, Habelschwerdter Allee 45, 14195 Berlin, Germany. Fax: þ49 30 838 556 20. E-mail addresses: benny.briesemeister@fu-berlin.de (B.B. Briesemeister), lars.kuchinke@rub.de (L. Kuchinke), ajacobs@zedat.fu-berlin.de (A.M. Jacobs). brain research 1564 (2014) 62–71