Psychological Medicine, 1999, 29, 161–170. Printed in the United Kingdom  1999 Cambridge University Press Increased automatic spreading activation in healthy subjects with elevated scores in a scale assessing schizophrenic language disturbances S. MORITZ, B.ANDRESEN, F.DOMIN, T.MARTIN, E.PROBSTHEIN, G.KRETSCHMER, M.KRAUSZ, D.NABER  M. SPITZER From the Universita  ts-Krankenhaus Hamburg-Eppendorf, Klinik fu  r Psychiatrie und Psychotherapie, Hamburg ; and Universita  tsklinikum Ulm, Abteilung Psychiatrie, Ulm, Germany ABSTRACT Background. Previous studies on semantic priming have suggested that schizophrenic patients with language disturbances demonstrate enhanced semantic and indirect semantic priming effects relative to controls. However, the interpretation of semantic priming studies in schizophrenic patients is obscured by methological problems and several artefacts (such as length of illness). We, therefore, used a psychometric high-risk approach to test whether healthy subjects reporting language disturbances resembling those of schizophrenics (as measured by the Frankfurt Complaint Questionnaire subscale ‘ language ’) display increased priming effects. In addition, the Schizotypal Personality Questionnaire was used to cover symptoms of schizotypal personality. Enhanced priming was expected to occur under conditions favouring automatic processes. Methods. One hundred and sixty healthy subjects performed a lexical decision semantic priming task containing two different stimulus onset asynchronicities (200 ms and 700 ms) with two experimental conditions (semantic priming and indirect semantic priming) each. Results. Analyses of variance revealed that the Frankfurt Complaint Questionnaire-‘ language ’ high scorers significantly differed from low scorers in three of the four priming conditions indicating increased automatic spreading activation. No significant results were obtained for the Schizotypal Personality Questionnaire total and subscales scores. Conclusions. In line with Maher and Spitzer it is suggested that increased automatic spreading activation underlies schizophrenia-typical language disturbances which in our study cannot be attributed to confounding variables such as different reaction time baselines, medication or length of illness. Finally, results confirm that the psychometric high-risk approach is an important tool for investigating issues relevant to schizophrenia. INTRODUCTION Over the past thirty years a growing body of evidence has been accumulated suggesting that semantic memory is based upon map-like structures in which concepts (represented as nodes) are placed according to semantic related- ness (see Collins & Loftus, 1975; den Heyer &  Address for correspondence : Steffen Moritz, Universita  ts- Krankenhaus Hamburg-Eppendorf, Klinik fu  r Psychiatrie und Psychotherapie, Martinistrasse 52, 20246 Hamburg, Germany. Briand, 1986). If a certain node (for example, the word ‘dog’) is activated, activation is thought to spread to adjacent (i.e. related) concepts (for example, ‘cat’ or ‘bone’) until it decays either by some kind of inhibitory process or because of fading strength. Activated nodes are considered to have lowered thresholds for items matching their identity so that processing of these items is facilitated once a node is activated (see Neely, 1991, for a review). Recent findings share the common assumption that the 161