20 Abstracts /International Journal of Psychophysiology 25 (1997) 17-84 variables such as word frequency, number of letters, syllables, phonology, orthography, syntax, focus of attention and in- stance-category distribution. The subjects were undergraduates of literature and lan- guage who performed a three choice task in a single expres- sion condition and in a context condition (the original poem). After target presentation, subjects decided whether the ex- pression was literal, metaphoric or anomalous, while reaction times and EEG were recorded. At Fz, the peak amplitude of the N400 was highest for anomalies, lowest for hterals, whereas metaphors took an intermediate position. Interactions of ex- pression type with condition revealed that the higher N400- amplitude for metaphors compared with literals decreased in context compared with the single expression condition. This interaction was invertedly related with the percentage of cor- rect decisions. Anomalies remained recognizable in all conditions, Meta- phors became more anamolous when they were not in (a poetic) context. Context mitigated recognizability for the liter- als, which became more metaphoric. Since metaphoricity was predicted by the context, uncertainty and unexpectedness were reduced, particularly at frontal locations. Thus, N400 had a ‘stepwise’ sensitivity for unexpectedness, while context dimin- ished its intensity. ERPS REFLECTING THE ON- AND OFFSET OF SELEC- TION PROCESSES Rob H.J. Van der Lubbe* and Jaap C. Woestenburg Free University Amsterdam, The Netherlands Visual attention has been studied from both a behavioural and a physiological perspective. Both perspectives (1) provide evidence for the view that attentional selection can occur before a stimulus has been identified. On the basis of a recent theory of visual attention (2) we distinguished three selection processes: location selection (automatic or controlled), object selection and attribute selection. An advantage of the physio- logical perspective is that ERP measures may provide infor- mation about the onset or offset of these selection processes. For instance, a reflection of the onset of object selection may be observed in a contralateral enhancement when targets are presented in the left or right hemifield. Bilateral six-element-arrays (five distracters and one out of two possible targets) were preceded by peripheral cures (SOA: 100-300 ms). The cues differed in size and indicated either: 1) the precise target position, 2) the target side, or 3) the complete array. The target ERPs were corrected for contribu- tion of the cue ERP by using the OPTA (3). The contralateral enhancement of a posterior P150 and N230 were interpreted as reflections of the onset of object selection and the offset of automatic (the P150) and controlled (the N230) location selec- tion. Ipsilateral activity of the N230 depended on the type of cue, which suggests that controlled location selection was affected by automatic selection. Modulation of P3 amplitude may be interpreted as the reflection of attribute selection. (1) Hillyard et al. (1994). In: Cognitive Electrophysiology, Birkhluser, Boston. (2) Van der Heijden, A. H. C. (1992). Selective Attention in vision, Routledge, New York. (3) Woestenburg et al. (1983). Biological Psychology, 17, 173-191. RELATIONSHIPS BETWEEN EEG ASYMMETRIES AND BILATERAL EDA AND POSSIBLE RELEVANCE FOR PSYCHOSOMATICS I. Papousek* and G. Schulter Dept. Psychology, University of Graz, Univ. Platz 2, A-8010 Graz, Austria In spite of a large number of studies on this issue, little is known about hemisphere influences on electrodermal activity (EDA) up to data. Instead, quite an equivocal picture of relationships between laterality measures and EDA is pro- vided from the literature. At present, for instance, no direct relationships between EEG asymmetries and spontaneous electrodermal fluctuations has been shown. In a sample of 60 subjects, EEG (Fpl, Fp2, F3, F4, T3, T4, P3, P4) and bilateral spontaneous EDA were measured at the same time. Aditio- nally, some personality and emotional state measures were obtained. The results provide evidence for a clear association between EDA (lability and asymmetry) and frontal EEG asymmetry. This relationship seems to be modified by person- ality variables. Additionally, changes in electrodermal asym- metry from one recording session to another one are corre- lated with changes in EEG asymmetry and with changes in subject’s emotional state. Cortical influences on EDA asymmetry seem to be predominantly contralateral excitatory. Finally, subject’s amount of physical complaints can be pre- dicted by electrodennal lability and asymmetry, but especially by the interaction of these two variables. The conclusion is drawn that the great relevance of autonomic nervous system measures in lateral&y research is not to assess cortical asymmetries directly (this can be better done by other mea- sures), but to offer the possibility to observe consequences of cortical asymmetries on the autonomic nervous system, which may be influenced by certain factors. This could be of special importance for psychosomatics. INTRACRANIAL MISMATCH NEGATIVITY AND ITS COMPUTER SIMULATION J.D. Kropotov’*, R. NBltlnenZ, A.V. Sevostianov’, K. Alho’, K. Reinikainen’, O.V. Kropotova’ and E. Schwartz3 ‘Institute of the Human Brain of Russian Academy of Sci- ences, St. Petersburg, Russia; ‘Cognitive Psychophysiology Research Unit, University of Helsinki, Helsinki, Finland; 3Laboratoty for Natural Computing, Radford University, USA The mismatch negativity (MMN) was recorded in 11 patients (4 Parkinsonian, 4 OCD and 3 epileptic) in whom gold elec- trodes were implanted into the target structures. Each patient was presented with a randomized sequence of three tones, a