BRAIN IMAGING NEUROREPORT 0959-4965 & Lippincott Williams & Wilkins Vol 12 No 16 16 November 2001 3633 Visual spatial localization con¯ict: an fMRI study Edward L Maclin, 1,CA Gabriele Gratton and Monica Fabiani Department of Psychology, University of Missouri-Columbia, Columbia MO 65211, USA 1 Present address: Beckman Institute, 405 North Mathews Avenue, Urbana, IL 61801, USA CA Corresponding Author Received 22 August 2001; accepted 7 September 2001 fMRI and ERP studies have shown that tasks comprising con¯icting stimulus±response associations activate a variety of cortical regions. It remains unclear whether any of these areas are activated by all con¯ict tasks, or whether con¯ict resolu- tion is a common property of a number of distinct anatomical regions. Several regions in frontal and parietal cortex are activated by both exogenous (position) and endogenous (arrow direction) localization cues. The present event-related fMRI study used a version of the Simon task with independent positional and directional cues. The results indicated that spatial localization con¯ict activated pre-motor and superior parietal regions in the right hemisphere known to be involved in spatial localization, but anterior cingulate activation did not reach threshold. This suggests that con¯ict within a single functional modality may be processed in the region embodying that modality, and anterior cingulate may be called on only to resolve con¯ict between modalities. NeuroReport 12:3633±3636 & 2001 Lippincott Williams & Wilkins. Key words: Anterior cingulate cortex; fMRI; Premotor cortex; Simon effect; Response con¯ict; Visual spatial localization INTRODUCTION In the original Simon experiments it was discovered that reaction times to lateralized auditory stimuli were faster with the ipsilateral hand than with the contralateral hand [1]. This was interpreted as re¯ecting the prepotency or intrinsically stronger stimulus±response association of ipsi- lateral responses and a potent stimulus±response incom- patibility generating con¯ict in the contralateral condition. Subsequent spatial response con¯ict studies have em- ployed a wide range of procedures, including Stroop [2], Eriksen [3] and Simon tasks [4]. The notion of prepotency is central to the design and interpretation of con¯ict studies [5]. Prepotency can arise in tasks where, as in the original Simon task, there is an apparent anatomical basis for the prepotency, as well as in tasks comparing linguistic and non-linguistic perception (as in the classic Stroop task), in wholly verbal tasks like verb generation [6], and in over- learned tasks where an arti®cial prepotency can be demon- strated [7]. Unfortunately, the concept of prepotency is as adaptable as it is important. The ¯exibility of the concept of prepotency is re¯ected in the wide range of experimental tasks and paradigms that have been used in studies of response con¯ict. This breadth of experimental paradigms makes it dif®cult to compare results across experiments, though it has led to much constructive debate concerning the dimensional overlap of stimuli and responses [8] and over how con¯ict resolution might be parsed neuroanatomically [9]. Numer- ous functional imaging studies have demonstrated activa- tion of discrete regions of medial frontal cortex, and in particular the anterior cingulate cortex, during response con¯ict tasks [10]. We are unaware, however, of any study that has looked at con¯icting spatial localization cues. The goal of the present study was to determine the pattern of cortical activation produced by a simple visual Simon task, which could be elaborated in future studies to speci®cally address some of the outstanding questions regarding con¯ict processing in the human brain. This task presents a con¯ict between two independent cues (location and direction) within the modality of visual-spatial locali- zation. These cues can be considered as exogenous and endogenous cues, respectively. A recent fMRI study by Rosen et al. [11] using these types of stimuli concluded that both cues activate parts of cingulate, supplementary motor, bilateral parietal, and dorsal premotor cortex, as well as thalamic and cerebellar regions. The task used in the present study is similar to one that has been used in a series of behavioral and ERP studies addressing the issue of response prepotency. In the initial study of this series, Eimer [12] showed that single arrow stimuli caused a lateralized potential whose polarity was determined by the direction pointed to by the arrow, even when the arrow did not predict the correct response hand, demonstrating intrinsic response mapping for arrows. Fournier et al. [3] also demonstrated the prepotency of arrows by comparing letters (S, H, C and N) to symbols (,, . , / and \) in a ¯anker compatibility task. For symbol stimuli the compatibility effect was greater than for the arrow stimuli; this was interpreted as evidence of an intrinsic response association for arrows. Most recently,