Psychological Research (2004) 68: 18–30 DOI 10.1007/s00426-002-0128-z

O R I GI N A L A R T IC L E

Edmund Wascher Æ M. Wolber

Attentional and intentional cueing in a Simon task: An EEG-based approach

Received: 24 July 2002 / Accepted: 11 November 2002 / Published online: 15 May 2003
Springer-Verlag 2003

Abstract Advance information about the location of a stimulus (attentional cueing) does not affect the Simon effect (a shortening of manual response times whenever the position of a stimulus that is irrelevant for the task corresponds to the side of the response). However, advance information about the side of a response (intentional cueing) enhances the Simon effect. At first sight, these well-established results contradict two important assumptions about the origin of the Simon effect: (a) the effect originates at least in part in a covert shift of visual attention that forces the preparation of a response towards the location of the attentional shift and (b) interference between stimulus location and response side takes place within a response selection stage. We replicated the behavioral finding in a study that measured event-related potentials (ERPs) of the EEG. ERPs indicated that the mechanisms causing the Simon effect remain widely unaffected by advance information. Clear evidence for both response preparation and attentional shifts in the cue–target interval was found. Additionally, ERPs suggested that the increment of the Simon effect by intentional cueing might be due to perceptual factors rather than to an alteration in the mechanisms involved in the generation of a regular Simon effect. The implications of these data for the role of attention and of response selection in Simon tasks are discussed. Keywords Simon effect Æ Spatial attention Æ ERP Æ LRP Æ N1 Æ N2pc Æ Response preparation

E. Wascher (&) Æ M. Wolber Max Planck Institute for Psychological Research, Cognitive Psychophysiology of Action, Amalienstr. 33, 80799 Munich, Germany E-mail: [email protected] E. Wascher Æ M. Wolber Department of Clinical and Physiological Psychology, University of Tuebingen

Introduction Manual responses are faster if the location of a stimulus corresponds with the side of the response (Fitts & Seeger, 1953). This effect of spatial stimulus–response compatibility appears even if the location of the stimulus is not task relevant (Simon & Rudell, 1967). Two main questions are discussed in connection with the origin of this so-called Simon effect. First, why and how is the irrelevant spatial information processed, and second, which mechanisms are responsible for the impact of irrelevant spatial information upon performance. Both questions had been investigated by tasks that used precues of either the location of the imperative stimulus (attentional precueing) or the side of the required response (intentional precueing). Unfortunately, despite the increasing number of studies published so far, the results obtained remain contradictory with respect to some important assumptions concerning the Simon effect. Mechanisms generating a spatial code In the processing of the irrelevant spatial information of the imperative stimulus (its location), two approaches have been presented. The attentional-shift hypothesis (Nicoletti & Umilta`, 1994; Rubichi, Nicoletti, Iani, & Umilta`, 1997; Stoffer, 1991; for a review see Stoffer & Umilta`, 1997; Stoffer & Yakin, 1994; Umilta` & Nicoletti, 1992) postulates that a spatial code is generated when there is a shift in spatial attention towards the location occupied by the imperative stimulus (i.e., the stimulus that delivers the information about how to respond). The alternative assumption, the referential coding account (Hommel, 1993), assumes that a spatial code is formed by relating the imperative stimulus to a reference frame. Although these two accounts are sometimes presented as opposing assumptions, there are only minor differences between them: both theories assume that spatial coding is not retinotopic and occurs

19

automatically, that spatial frames are variable and initially anchored at a reference object defined either by intention (referential coding) or by the focus of attention (Stoffer & Umilta`, 1997). The main contradictions between the two accounts concern the definition of the origin of the reference system. The attentional-shift account postulates that the origin of the reference system that determines the direction and amount of the Simon effect is the focus of attention just before the imperative stimulus is presented. Only the last shift of spatial attention preceding the presentation of the imperative stimulus is responsible for the direction of the Simon effect (Nicoletti & Umilta`, 1994; Rubichi et al., 1997; Stoffer, 1991). The predictions for attentional cueing experiments are clear-cut: a spatial cue that predicts the most probable location of the target stimulus should evoke a shift of attention towards the cued location. Thus, if the imperative stimulus appears at the cued location, the Simon effect should disappear because no more attentional shift is necessary towards its location. The referential coding account would assume that despite the attentional cue the reference frame remains the same and therefore attentional cueing should not affect the Simon effect. In accordance with the attentional-shift account, Stoffer & Yakin, (1994) report a decrease in the Simon effect with peripheral precues. In 75% of trials, Stoffer & Yakin, (1994) presented a precue at a peripheral location that validly predicted the location of the target stimulus. The Simon effect in attentionally cued trials was compared to neutrally cued trials that started with a central precue. The Simon effect was markedly smaller with validly cued trials compared to neutrally cued trials. In another experiment, Umilta` & Liotti, (1987) reported that for any irrelevant spatial parameter that is validly given in advance, the Simon effect can no longer be observed. They presented target stimuli at one out of four possible stimulus locations, two to the left and two to the right of a fixation cross. Responses could correspond with respect to the absolute location of the stimulus and with respect to its relative location within the left or right pair. Presenting valid information about either the side of the display or the relative location of the target eliminated the Simon effect associated with the cued spatial parameter but not with the other one. In contrast to these findings, Zimba & Brito, (1995) report a persisting Simon effect with peripheral precues within a wide range of stimulus onset asynchronies (SOAs) (up to 1000 ms). In contrast to Stoffer & Yakin, (1994) and to Umilta` & Liotti, (1987), Zimba & Brito, (1995; experiment 2) did not present neutral precues but invalid precues in 20% of the trials. Thus, the main difference between tasks that show a reduction of the Simon effect in accordance with the attentional-shift account and those that show remaining Simon effects was the validity of cue information. Studies in which the cue information was 100% valid with respect to the stimulus location (Stoffer & Yakin, 1994; Umilta` & Liotti, 1987; Van der Lubbe & Woestenburg, 1999) show

a reduction in the Simon effect, whereas studies that used cue validities lower than 100% (e.g., Zimba & Brito, 1995) showed no change of the Simon effect due to attentional precueing. Additionally, a number of studies that used symbolic, centrally presented precues replicate the latter finding (Proctor, Lu, & Van Zandt, 1992; Verfaellie, Bow, Bowers, & Heilman, 1988; Zimba & Brito, 1995). Thus, some of the results obtained so far in attentional cueing paradigms on the Simon effect contradict the attentional-shift account proposed by Umlita` and co-workers (Nicoletti & Umilta`, 1994; Umilta` & Nicoletti, 1992). On the other hand, the referential coding account cannot conclusively explain numerous findings on the Simon effect. First of all, it has been argued that the theory is very flexible with respect to the reference object chosen by the subject (Stoffer & Umilta`, 1997). In attentional cueing it is unclear why the center of the reference system remains between the two possible stimulus positions with 80% valid intentional cueing, whereas it moves to the target position if the validity of the cue increases to 100% (Stoffer & Yakin, 1994; Van der Lubbe & Woestenburg, 1999). Second, the referential coding account cannot explain the sensitivity of the Simon effect to particular manipulations of the stimulus material: (1) The Simon effect decreases with the existence of a fixation cross (Proctor & Lu, 1994). With respect to the referential coding account, the fixation cross should provide a useful reference point that should enhance the Simon effect, whereas attentional shifts might be facilitated by the lack of an attentional anchor (Yantis & Jonides, 1990). (2) In a cueing experiment where the centrally presented cue delivers information about whether it was a go or a no-go trial, the direction of the Simon effect depended on the duration of stimulus presentation, although all visual elements of the stimulus material that might have been the basis of the reference system remained the same (Rubichi et al., 1997). (3) The Simon effect increases if a noise stimulus that accompanies the target stimulus has the same color as the target stimulus compared to differing colors (Proctor & Lu, 1994). There is no convincing argument explaining why the reference frame should change with the relation of colors between two elements of a visual display. Interaction of stimulus and response codes Most theories on the Simon effect assume that the interference between the spatial code of the stimulus and the response takes place in a response-selection stage (e.g., Lu & Proctor, 1995; Proctor et al., 1992; Umilta` & Nicoletti, 1990). The spatial code that is obtained from processing the irrelevant stimulus location is assumed to activate a corresponding response via an unconditional (direct, automatic) route (De Jong, Liang, & Lauber, 1994; Kornblum, Hasbroucq, & Osman, 1990), whereas the relevant stimulus dimension (e.g., letter identity, shape, or color) is processed in a conditional (indirect,

20

controlled) manner. When the two routes are generating the same code (S–R corresponding condition), the correct response can be selected very fast. If the two routes activate different response codes, the correct response has to be selected against incorrect activation from the unconditional route. This takes time and therefore responses are delayed. Recent evidence in favor of automatic activation of the motor system comes from studies on the Simon effect that measured the lateralized readiness potential (LRP). The LRP is an EEG measure that is based on the increased activity of the motor cortex in the hemisphere contralateral to a planned or executed response (De Jong, Weirda, Mulder, & Mulder, 1988; Gratton, Coles, Sirevaag, Eriksen, & Donchin, 1988). The LRP (for more details see below) shows that EEG activity over the motor cortex is initially increased contralateral to the position of the laterally presented stimulus (Valle-Incla´n, 1996; Wascher & Wauschkuhn, 1996). This early component of the LRP is assumed to reflect direct visuomotor activation evoked by the stimulus. This notion is supported by the predictive value of the LRP amplitude for the amount of the Simon effect. In patients who suffer from Parkinson’s disease, the early LRP component is markedly enlarged if compared to healthy controls. These patients show a larger Simon effect, especially after non-corresponding trials and commit more errors. In both groups, non-corresponding trials in which a large early LRP component was evoked were erroneously responded to (Praamstra & Plat, 2001). So far, response activation is supposed either to support (in corresponding trials) or to hamper (in non-corresponding trials) the selection of the required response. Some studies intended to verify the response selection account by the use of intentional precueing. In these studies the most probable response was validly given in advance (mostly 80%). However, the predictions of this method regarding the Simon effect are unclear. Assuming that the intentional precue leads to a preselection of the upcoming response, an effect based on response selection should be markedly reduced. Hitherto, however, experimental findings regarding intentional precueing and the Simon effect show a consistently contradictory pattern. Verfaellie (Verfaellie et al., 1988) and Proctor (Proctor et al., 1992; Proctor & Wang, 1997) report that the Simon effect did not decrease but even increased with valid intentional precues. Moreover, even when the side of response was given 100% valid in advance and the task was a go–no-go task, a reliable Simon effect was obtained (Hommel, 1996). On the other hand, one might argue that the precue solely activates an additional spatial code. In this case, valid cues do not necessarily reduce the Simon effect. Modulations of the Simon effect should depend on the way the three spatial codes (cued response, stimulus location, and required response) interact. However, so far no model has been discussed that integrates the additional code into common models concerning the Simon effect. The simplest approach—to argue that

the speed of response selection depends only on the existence of an interference between codes—does not fit the data reported above, because in this case no Simon effect should be obtained in invalidly cued trials. Thus, behavioral data obtained with both attentional and intentional cueing apparently do not support common assumptions concerning the Simon effect. But are these findings really sufficient to reject repeatedly confirmed assumptions? Rather than doubting these assumptions one might argue that the subject did not perform according to the instruction, possibly due to the uncertainty of the advance information (note that the contradicting results for attentional cueing are restricted to tasks where cue validity was below 100%). Additionally, one might argue that the task had changed due to advance information and therefore the cueing task might be of no informational value for the regular Simon effect obtained with no cues. Both assumptions can be tested by the use of event-related EEG activity. First, attention- and motor-related EEG activity can be measured in the two tasks (attentional and intentional cueing) to obtain a measure of task-specific processing, and second, specific correlates of processing in the Simon tasks (such as the LRP) can be compared to a regular, uncued Simon task. EEG correlates of visual attention Studies on event-related potentials of the EEG (ERPs) have reported primarily three components related to spatial attention. Two early sensory-evoked components (N1 and P1) vary depending on whether a stimulus is presented at an attended or an unattended spatial location (Eason, 1981; Hillyard & Mu¨nte, 1984; Mangun & Hillyard, 1987): P1 amplitude is reduced whenever a stimulus is presented at an unattended location but does not differ between stimuli at an attended position or stimuli presented after neutral cues. P1 effects are not observed if also invalidly cued stimuli have to be processed (Eimer, 1994). The subsequent N1 is enlarged for stimuli presented at an attended location compared both to unattended or neutrally cued stimuli (Luck, 1995; Luck et al., 1994). Besides these two ERP effects, attention-related ERP components were also observed in tasks where the location of attention was not defined by external information given either by instruction (sustained attention) or by spatially informative cues (trial by trial). In visual search tasks, an increase in the N2 component at posterior electrode sites contralateral to the location of a pop-out stimulus was observed (N2pc; Luck & Hillyard, 1994a, 1994b). This component is only visible if a relevant element (either target or non-target) in a stimulus array has to be evaluated (Luck & Hillyard, 1994b), which might reflect that it is necessary to suppress irrelevant information surrounding a visual stimulus that has to be identified (Luck, 1995; Luck & Hillyard, 1994b). However, some studies have also

21

found an N2pc in tasks where the target was accompanied by a single distractor item only (Eimer, 1996; Wascher & Wauschkuhn, 1996). Therefore, it might more generally reflect the allocation of attention at a relevant position within a multiple visual display. If no allocation of a target among distractors is necessary, as for example for isolated targets, no N2pc can be observed (Luck, 1995; Luck & Hillyard, 1994b). Also, with attentional cues, the localization of validly cued targets is facilitated and therefore no, or at least a reduced N2pc should be expected.

areas, which accompanies shifts of visual attention but is not directly caused by it. Thus, if the processing in a Simon task changes by manipulation of the experimental set-up by attentional or intentional cueing, the early deflections in the LRP should change correspondingly.

Experiment 1. Attentional cueing Methods Participants

EEG correlates of movement preparation and sensori-motor integration Response preparation is reflected in the lateral readiness potential (LRP). The LRP bases on increased activity of the motor cortex contralateral to the prepared limb prior to any voluntary movement (Kornhuber & Deecke, 1965). By processing a difference potential between activity over the contra- and ipsilateral hemispheres and averaging these difference waves (De Jong et al., 1988; Gratton et al., 1988), the time course of movement preparation can be observed independently from behavior. If a centrally presented cue delivers information on the subsequent response, an LRP-like asymmetry was reported with a peak latency between 250 and 400 ms that was initially interpreted as automatic response activation evoked by the cue (Eimer, 1995). More recently, Verleger, Vollmer, Wauschkuhn, Van der Lubbe, & Wascher, (2000) pointed out that the maximum of this component was located not directly over the hand motor area but over more anterior areas. Additionally, this component is visible even if no manual but a saccadic response is required. Thus Verleger and coworkers, (2000) argue that this component might reflect action-related spatial processing in premotor areas. In cueing tasks, a subsequent sustained LRP component at central sites was found to be specific to the preparation of a finger movement (Van der Lubbe et al., 2000). As already mentioned, the LRP additionally indicates response activation evoked by the laterally presented visual stimulus in a Simon task. Following the presentation of a lateral target stimulus, a phasic component over the motor cortex indicates the activation of a response that corresponds to stimulus location. Studies that investigated cortical asymmetries not only over the motor cortex revealed a high temporal coincidence of this initial motor component and the N2pc (Valle-Incla´n, 1996; Wascher & Wauschkuhn, 1996). Based on the finding that the amplitudes of anterior and posterior asymmetries within this time window vary independently from each other (e.g., Praamstra & Plat, 2001; Wascher, Schatz, Kuder, & Verleger, 2001), Oostenfeld and co-workers (Oostenveld, Praamstra, Stegeman, & van Oosterom, 2001) concluded that the anterior manifestation of an N2pc-like component might reflect attention-related activation of premotor

Fourteen female students of the University of Tuebingen, aged between 18 and 35 years (mean age, 23.6), were recruited for this experiment. All participants were right-handed, in good physical health, and had no history of psychiatric or neurological disorder. They had normal or corrected-to-normal vision. Participants received either course credit or were paid about 30 DM (approximately US $15) for participation in this experiment. Stimuli and procedure Participants were seated in a comfortable armchair in a soundproof, electrically shielded chamber. Visual stimuli were presented on a 22-inch Multisync monitor with an observation distance of approximately 1.2 m. An IBM-compatible PC controlled the presentation of the task. A white fixation cross in the center of the screen and two symmetrically positioned white frames (representing the possible stimulus positions) were displayed continuously. Each trial started with an arrow (visual angle, approximately 0.7) that was presented for 200 ms. The arrow pointed either to the left or to the right. It was located in the center of the screen, directly on the fixation cross. The arrow indicated the side of the relevant stimulus validly for 80% of the trials. After a variable delay between 450 and 550 ms (mean delay, 500 ms) a letter (A or B) appeared within one of the two lateral frames and a noise stimulus (three horizontal bars, similar to the letters in size and luminance) appeared simultaneously at the opposite location. Symmetrical noise stimuli were introduced primarily to avoid exogenous, purely sensory driven asymmetries in the EEG (see Praamstra & Plat, 2001; Wascher & Wauschkuhn, 1996). These additional stimuli do not change the characteristics of the effect essentially (Wascher et al., 2001) and may even enhance the behavioral effect (Proctor & Lu, 1994). Target stimuli were also presented for 200 ms. The distance of the inner border of the two lateral frames from the fixation cross was 1.1. Stimuli were approximately 11 mm wide and 13 mm high (0.5 · 0.6) and of bright yellow color. The four types of stimuli (left A, right A, left B, right B), valid and invalid cues, were presented in randomized order. The intervals between an imperative signal and the presentation of the next cue varied between 1700 and 2700 ms (on average 2200 ms), with 750 stimuli presented (600 validly, 150 invalidly cued). The duration of the task was about 35 min. Two breaks were set by the experimenter between trials 200 and 300 and between trials 450 and 550. In the same session, either before or after the cueing task (counterbalanced across subjects), subjects performed a Simon task without cues. Only the imperative stimuli from the cueing task were presented with an inter-trial-interval between 1700 and 2700 ms (on average 2200 ms). Subjects performed 400 trials in this task.

Recording and data processing An EEG was recorded from 51 scalp positions covering the entire scalp using Ag/AgCl electrodes (Picker-Schwarzer), with an

22 electrode affixed at the nose as reference. For control of ocular artifacts, the electro-oculogram (EOG) was recorded both vertically from above vs below the left eye (vEOG) and horizontally from the outer canthi of both eyes (hEOG). EEG and EOG were amplified and filtered by Psylab amplifiers with 5.31 s the time constant (0.03–70 Hz bandpass). Response force was recorded continuously from nearly isometric weight elements, which had to be pressed by the two index fingers. Criterion for a given response was a force of 2 N (N)1. Triggered by signals from the control computer (IBM compatible), the data (EEG, EOG, and response force) were digitized at 200 Hz (every 5 ms) and stored on another PC (IBM compatible), from 100 ms before to1900 ms after onset of the cueing stimulus in the cueing task. In the task without cues, data were digitized at 500 Hz (every 2 ms). All other parameters were the same. Trials with zero lines, out-of-scale values, slow drifts larger than 80 lV and fast shifts larger than 100 lV/500 ms were excluded from further analyses. Trials affected by ocular potentials were not excluded but the transmission of vEOG and of hEOG into the EEG was estimated by a regression method (see Verleger, Gasser, & Mo¨cks, 1982) and was subtracted from the EEG data.

ANOVAs including the factors validity (2)3 and S–R correspondence (2). Additionally, the amplitudes and latencies of the N1 at posterior electrodes (PO7/PO8) contra- and ipsilaterally to the target stimulus were analyzed in an ANOVA with the factors validity (2) S–R correspondence (2) and electrode position (2, contralateral vs ipsilateral). (b) Onset latency and amplitude of the N2pc (as indicated in the ERL) were measured at parieto-occipital sites. Onset latencies were measured by the Jack-Knife Procedure proposed by Miller and co-workers (Miller, Patterson, & Ulrich, 1997; Ulrich & Miller, 2001) using a 50% amplitude criterion. They were analyzed in an ANOVA with the factors validity (2) and S–R correspondence (2). The amplitude of the N2pc was measured as the mean amplitude in the difference wave at PO7/PO8 around the maximum (±40 ms) of asymmetry as indicated by the grand average for all three cueing conditions (valid, invalid, no cue). Note that for the analyses the ERLs were inverted for S–R non-corresponding trials because we were interested in the amount of asymmetry evoked by the target stimulus. Data from this measure were entered into an ANOVA with the factors cueing (3, valid vs invalid vs no cue) and S–R correspondence (2). Finally (c), asymmetries over the motor cortex (LRP; C1/C24) were analyzed within the same time window and by the same procedure as the amplitude of the N2pc.

Data analysis Response parameters Response time was defined as the moment when response force crossed the criterion of 2 N. Trials with incorrect responses (forces larger than 2 N on the incorrect hand) were defined as response errors and excluded from response time and EEG analyses. Mean response times for each participant were entered to an ANOVA with the independent variables S-R correspondence (2, corresponding vs non-corresponding trials) and cue validity (3, valid vs invalid vs no-cue). Subsequently, posttests of response times comparing the two validity conditions and either condition in the cued task with the regular Simon tasks were calculated. Error percentages were entered to the same analyses as response times. Effects with df >1 in the numerator were corrected by Greenhouse-Geisser Epsilon (e). EEG parameters Event-related potentials and event-related lateralizations (ERLs)2 were calculated for trials without response errors, both time locked to the cue (S1) and time locked to the target stimulus (S2). ERLs during the S1–S2 interval were analyzed in an average potential across all conditions since participants could not know in advance whether the arrow indicated the stimulus location validly or not, whether stimulus and response side would correspond or not. During the S1–S2 interval, whether the cue activated a spatial code and whether subjects prepared a manual response were both tested. Central asymmetries at C1/C2 (both an early phasic component reflecting the processing of the cue and a sustained activity in the intentional cueing task were well defined at this electrode site) were measured as mean amplitudes between 250 and 350 ms and during the last 100 ms preceding the imperative stimulus. Mean amplitudes within these time windows were tested by a t-test against zero. In the EEG activity evoked by the target stimulus, the following components were analyzed: (a) The amplitudes and latencies of the posterior N1 (always measured as the most negative peak between 150 and 250 ms) were analyzed at Pz, POz and Oz in separate 1

Newton is the unit of force. 2 N corresponds to the weight exerted by a mass of 204 g. This criterion is well within the range used by usual (all-or-none) response devices. 2 ERLs are difference potentials calculated as the LRP (Coles, 1989; Wascher and Wauschkuhn, 1996) that indicate spatial processing. Contralateral-ipsilateral differences are calculated for all symmetrical electrode pairs by subtracting the EEG activity ipsilateral to the required response from the contralateral activity. Mean differences are calculated separately for left and right responses and subsequently averaged.

Results Response parameters Response times (see Table 1) were only marginally affected by the cueing condition (F(2,26)=3.05, MSE=1623.63, e=.81, p=.078). None of the two cueing conditions differed significantly from the regular effect without cues (valid vs no cue: F(1,13)=2.58, MSE= 2227.62, p=.132; invalid vs no cue: F(1,13)<1). A separate analysis of the cueing task, however, showed an acceleration of responses with valid cues if compared to invalid cues (F(1,13)=8.92, MSE= 883.34, p=.011). An effect of S–R correspondence was observed (F(1,13)=48.88, MSE=267.72, p<.001) that did not interact with cue validity (F(2,26)<1). Response errors showed basically the same effects as response times except that there was no main effect of cue validity in the global analysis (F(2,26)=1.66, MSE=43.47, e=.56, p>.2). A separate analysis in the cueing task revealed a slightly better performance in validly cued trials (F(1,13)=5.15, MSE=13.34, p=.041). The effect of S–R correspondence (F(1,13)=24.97, MSE=9.74, p<.001) again did not 3 All analyses including temporal measures or amplitudes that were derived from peak picking are compared for the two cued conditions only because of the differing EEG-sampling rates in the two experiments. 4 C1/C2 was selected for the measurement of motor activation since it showed hand-motor activation most purely in a regular 10/ 10 system. This electrode pair is located over the hand motor areas and close to the electrode positions C3’/C4’ as defined by Kutas Donchin (1980). C3 and C4, which might be an alternative electrode pair to measure the LRP, are in many subjects located behind the central sulcus (see Steinmetz et al., 1989) and therefore more probably affected by the overlap of activity from the parietal cortex (see also Van der Lubbe et al., 2000; Wascher et al., 2001), which might interfere with motor activity if laterally presented stimuli are used.

23 Table 1 Response times and error rates for experiment 1 (stimulus cueing) and experiment 2 (response cueing) Cueing Experiment 1 Valid Invalid No cueing Experiment 2 Valid Invalid No cueing

S–R corresponding

S–R noncorresponding

Simon effect

458.5 (4.1) 478.8 (4.9) 470.9 (4.0)

479.9 (6.3) 507.0 (9.9) 496.2 (7.1)

21.4 (2.2) 28.2 (5.0) 25.3 (2.1)

422.9 (4.3) 469.7 (7.3) 459.9 (5.8)

460.6 (8.7) 477.8 (11.5) 483.5 (7.6)

37.7 (4.4) 8.1 (4.2) 23.6 (2.2)

vary with cue validity (F(2,26)=2.18, MSE=10.40, e=.66, p=.155). EEG parameters Central asymmetries evoked by the cue differed from zero during an early phasic component (t(13)=4.181, p=.001) but not preceding the target stimulus (t(13)=.564, p>.2; see Fig. 3, thin line and empty bars). N1 latency was not affected by cue validity at central electrodes (for all electrodes F(1,13)<1). Additionally, no change of N1 amplitude was observed with the validity of the cue (for all electrodes F(1,13)<1). At lateral electrodes a small but significant decrease in N1 latencies at electrode sites contralateral to the imperative stimulus was observed (195.5 vs 199.5 ms; F(1,13)=4.89, MSE=88.39, p=.046). This effect, however, was only found for invalidly cued trials (interaction electrode site by cue validity: F(1,13)=6.92, MSE=80.63, p=.021). Again no main effect of cue validity for N1 latencies was found (F(1,13)<1). N1 amplitudes were larger contralateral to the target stimulus (F(1,13)=17.37, MSE=296.10, p=.001). Amplitude effects did not interact with neither cue validity (F(1,13)=1.17, MSE=55.22, p>.2) nor was a main effect of cue validity observed (F(1,13)<1). Thus, early sensory components of the ERP seem to be widely unaffected by attentional cueing. However, N1 reflects only one distinct peak at the beginning of asymmetric activity that becomes obvious when contralateral-ipsilateral difference waves (event-related lateralizations, ERLs) are calculated (see Fig. 2). If looking at the latency range of the N1 (N1 peak latency in the ERP grand averages ranged between 185 and 200 ms) more closely, ERLs indicate that this measure might reflect different processes in validly and invalidly cued trials. In validly cued trials, the first peak of asymmetry appears around 180 ms, before the N1. In invalidly cued trials, the asymmetry starts only shortly before the N1 peak, around 175 ms. However, around the peak latency of the N1 (compare Fig. 1 and Fig. 2), the amplitude of the asymmetry was almost the same despite this large difference in temporal characteristics of lateralized activity. Measuring onset latencies of the ERLs (using a 50% amplitude criterion), a main effect of cue validity was

Fig. 1 ERPs recorded in experiment 1 (attentional cueing) at lateral and central parieto-occipital sites superposed for validly (bold lines) and invalidly cued (thin lines) trials. In the left column, recordings from parieto-occipital sites ipsilateral to the position of the target stimulus are depicted, in the right column, contralateral recordings. At neither electrode site was an effect of validity observed for the N1. N1 amplitude, however, was markedly enlarged at electrode sites contralateral to the target stimulus (for better comparison of contra- and ipsilateral amplitudes, a fine dotted line is added at the level of the amplitude at ipsilateral sites)

Fig. 2 Event-related lateralizations (ERLs) recorded in experiment 1 (attentional cueing). ERLs at a posterior electrode pair (PO7/ PO8) are depicted in the left and the middle panel for S–R corresponding (bold lines) and S–R non-corresponding (thin lines) trials. The peak in asymmetry is equivalent to the N2pc. In the left panel ERLs are depicted stimulus coded to highlight differences in onset-latencies: posterior asymmetries started earlier with valid attentional cueing (the thin vertical line indicates the onset of the N2pc in invalidly cued trials). Posterior ERLs In the middle panel and the LRP (right panel) are depicted response coded. The amplitude at posterior sites was enlarged with invalid attentional cueing (depicted in the second row). The LRP (measured at C1/ C2—close to the primary motor cortex—in the same time window as the N2pc) did not change with the type of cueing

observed (159.0 ms vs 200.4 ms; F(1,13)=11.561, p=.005), indicating that the posterior asymmetry started earlier in validly cued trials. S–R correspondence did not

24

affect the onset latency of the ERLs, and no interaction of cue validity by S–R correspondence was observed for this measure (both F(1,13)<1). The amplitude of the N2pc varied with the cueing condition (F(2,26)=4.42, MSE=28121.23, e=.67, p=.040). While it did not differ between the no-cue and the valid cue condition (F(1,13)<1), the amplitude of the ERL in the invalid cue condition was at least marginally increased both if compared to the valid cue condition (F(1,13)=4.39, MSE=60832.43, p=.056)) and if compared to the no-cue condition (F(1,13)=6.21, MSE=37634.48, p=.027). In contrast to the N2pc, the amplitude of asymmetries over the motor cortex did not change with the cueing condition (F(2,26)=2.16, MSE=1545.91, e=.81, p=.142). Thus, sensory and motor-related asymmetries in the EEG turned out to be at least in parts independent also in the present experiment. Discussion Behavioral data in the present study replicated previous studies on attentional cueing of the Simon effect: the effect of S–R correspondence on response time remained unaffected by the advance information about the most probable location of the imperative stimulus. For both validly and invalidly cued trials the Simon effect did not differ from a control condition where the stimulus location was not cued (denoted as the regular Simon effect). One might argue that the subjects did not perform according to instruction and did not shift their attention towards the position initially cued. This notion can be emphasized by the missing effect of validity on the N1. However, looking at the literature published so far on N1 amplitude effects with cue validity, the task usually used differs markedly from the present one. In those studies, subjects had to attend to a spatial position and decide whether a unilateral stimulus presented at that particular position was a target or not. Thus, in addition to the localization of attention, the attended and unattended positions differ with respect to the need to discriminate visual information. N1 effects related to spatial attention therefore might at least in part be due to the sensitivity of the N1 to stimulus discrimination (Vogel & Luck, 2000). Thus, the missing of the N1 effect in the present study is most probably due to the fact that on both validly and invalidly cued trials discrimination was necessary (see also Eimer, 1994). All other measures indicated clearly that a shift in attention had occurred after the attentional cue. (a) There was an overall effect of cue validity on response times with faster response times if the location of the relevant stimulus was validly cued. (b) ERL components related to shifts in spatial attention and/or stimulus localization (N2pc) started earlier in validly cued trials, and finally (c), the amplitude of the N2pc was enhanced in invalidly cued trials, indicating additional attentional processing in this particular condition.

Thus, the present data contradict the attentional-shift hypothesis in its current form. A shift in attention towards the location of the relevant stimulus prior to stimulus presentation did not affect the Simon effect. The same behavioral effect was observed for target stimuli at the attended location, for stimuli presented opposite the attended location, and in a task where no cues were presented. Additionally, while posterior EEG asymmetries indicated that attentional processing was altered by the type of cue used, central EEG asymmetries remained the same across all tasks and conditions, indicating a constant level of motor activation evoked by the laterally presented cue.

Experiment 2. Intentional cueing Response activation due to processing the irrelevant position of the target stimulus is generally assumed to affect a response selection stage, i.e., the response corresponding to the code delivered by the stimulus gets automatically activated. In case of correspondence between the preselected and the finally required response, response selection can pass a short route to response execution. In case of non-correspondence, however, the required response has to be selected against the activated one (Kornblum et al., 1990). The main argument appropriate to considering response selection as the relevant stage for effects of spatial S–R correspondence is the lack of a comparable effect in simple response tasks. In those tasks, S–R correspondence results in a response time effect of about 4–6 ms (Bisiacchi et al., 1994; Marzi, Bisiacchi, & Nicoletti, 1991), whereas the Simon effect is hardly ever smaller than 20 ms. As an alternative to the response selection account, Hasbroucq & Guiard, (1991) argued that the Simon effect might be caused by an interference of contradicting codes in stimulus identification rather than a response conflict. The distinction between these two accounts was explored by previous studies on intentional precueing. ‘‘By precuing either stimulus or response location, it should be possible to obtain evidence about whether the Simon effect has its basis in stimulusidentification or response-selection processes. [...] If the Simon effect is due to response-selection processes, its magnitude should also be influenced by advance information of response location’’ (Proctor et al., 1992, p. 56). Proctor et al., (1992) did not make any predictions about the direction of this influence and argue at the end of their study that the enhancement of the Simon effect they observed ‘‘implies that the Simon effect arises from response-selection processes rather than stimulus-identification processes’’ (Proctor et al., 1992, p. 73). A major concern with this argumentation was formulated by Buckolz, Odonnell, & McAuliff, (1996). ‘‘(...), if the Simon effect is due to a response-selection associated with the compatible output, this a posteriori retrieval advantage should be reduced or elimi-

25

nated when the required response turns out to be the precued one (...), since response retrieval has already occurred by the time the stimulus is delivered’’ (Buckolz et al., 1996, p. 550). In the following, Buckolz et al., (1996) discuss whether the task might have changed due to precueing and that consequently the cueing procedure might not be the adequate instrument to test mechanisms involved in a regular Simon effect. It is also possible that subjects do not preselect the cued response but activate a response code that is not directly related to advance response preparation if the validity of the intentional cue is below 100%. In this case, this spatial code may subjoin additional interference to the response selection process, leading to an increased Simon effect with intentional precueing. Taking these possibilities into account, the main goal of the present experiment was to explore whether subjects prepare the indicated response in the cue–target interval and whether ERP measures indicate an alteration of the task by the precueing procedure. Methods Participants Fourteen female students of the University of Tuebingen, aged between 18 and 33 years (mean age, 23.6), were recruited for experiment 2. None of the participants had taken part in experiment 1. They fulfilled the same criteria as the participants in experiment 1. Stimuli and procedure Stimuli and procedures were the same as in experiment 1 except that the arrow did not indicate the most probable location of the imperative stimulus but the side of the most probable response (validity 80%). Recording and data processing Recording and data processing were the same as in experiment 1. Data Analysis Data analysis was the same as in experiment 1.

Results Response parameters Response times were strongly affected by the type of cue used (F(2,26)=7.65, MSE=1440.79, e=.82, p=.005). This effect was based on an acceleration of responses if the cue validly indicated the side of the response (valid vs invalid: F(1,13)= 18.61, MSE=770.26, p=.001; valid vs no cue: F(1,13)= 7.36, MSE=1710.77, p=.018), while responses in invalidly cued and in the standard Simon task did not differ from one another (F(1,13)<1).

The effect of S–R correspondence (F(1,13)= 19.48, MSE=577.15, p=.001) additionally interacted with the type of the cue (F(2,26)=13.07, MSE=174.06, e=.67, p=.001). The amount of S–R correspondence differed between all cueing conditions (see Table 1; valid–invalid: F(1,13)=15.42, MSE=196.74, p=.002; valid–no cue: F(1,13)=10.97, MSE=62.69, p=.006; invalid vs no cue: (F(1,13)=9.37, MSE=88.97, p=.009). The largest Simon effect was obtained with valid cueing the smallest with invalid intentional cueing. Effects on response errors showed a qualitatively different pattern if compared to response times. Only a marginal effect of the type of cueing was observed (F(2,26)=3.69, MSE=30.97, e=.64, p=.065) that was primarily due to an increase in errors in invalidly cued trials if compared to validly cued trials (see Table 1; valid vs invalid: F(1,13)=6.78, MSE=17.28, p=.022; valid vs no cue: F(1,13)<1; invalid vs no cue: (F(1,13)=2.96, MSE=33.91, p=.109). The increase in errors with non-corresponding S–R relations (F(1,13)=19.23, MSE=13.38, p=.001) interacted also only marginally with the type of cueing (F(2,26)=3.67, MSE=4.88, e=.80, p=.052). Postanalysis revealed a reduced effect of S–R correspondence on errors in the standard, no cue condition if compared to the cueing task but no differences between valid and invalid cueing (see Table 1; valid vs invalid: F(1,13)<1; valid vs no cue: F(1,13)=11.20, MSE=2.07, p=.005; invalid vs no cue: (F(1,13)=3.57, MSE=5.45, p=.081). EEG parameters Both the first component of central asymmetry in the S1–S2 interval (t(13)=4.666, p<.001) and the central asymmetry preceding the target stimulus (t(13)=2.355, p=.035) differed from zero, indicating spatial code activation as well as ongoing movement preparation due to the information delivered by the cue (see Fig. 3). As in experiment 1, N1 latency was not affected by cue validity at central electrodes (Pz: F(1,13)=1.96, MSE=110.30, p=.185; POz: (F(1,13)<1; Oz: F(1,13)=<1). Additionally, no change in N1 amplitude was observed with the validity of the cue (for all electrodes F(1,13)<1). At lateral electrodes, N1 peaked earlier contralateral to the target stimulus (195 vs 199 ms; F(1,13)=6.38, MSE=64.65, p=.025). This effect was most pronounced when the cue did not point towards the subsequent target (validity by S–R correspondence by electrode position: F(1,13)=14.52, MSE=36.90, p=.002). Also, N1 amplitudes were enlarged contralateral to the position of the target stimulus (F(1,13)=19.70, MSE=399.64, p=.001), but did not differ between validly and invalidly cued trials (F(1,13)=1.52, MSE=297.72, p>.2). As in experiment 1, N1 reflected only one distinct peak at the beginning of asymmetric activity (see Fig. 4). More insight into asymmetric cortical activity is obtained when ERLs are calculated (see Fig. 5). The main difference with experiment 1 in terms of the temporal characteristics of ERLs appears to be that the

26

Fig. 3 Waveshapes (upper panel) and mean amplitudes (and standard error of means; lower panel) of central asymmetries in the S1–S2 interval. Data are depicted for experiment 1 (attentional cueing: thin lines, white bars) and experiment 2 (intentional cueing: bold lines, black bars). An early, phasic ERL component (between 250 and 350 ms) is visible in both experiments. Preceding the imperative stimulus, sustained EEG asymmetry indicating ongoing response preparation is visible with intentional cueing only (**: p<=.001; *: p<.05)

Fig. 4 ERPs recorded in experiment 2 (intentional cueing) at lateral and central parieto-occipital sites superposed for validly (bold lines) and invalidly cued (thin lines) trials. In the left column, recordings from parieto-occipital sites ipsilateral to the position of the target stimulus are depicted, in the right column, contralateral recordings. At neither electrode site was an effect of validity observed for the N1. N1 amplitude was again markedly enlarged at electrode sites contralateral to the target stimulus (for better comparison of contra- and ipsilateral amplitudes, a fine dotted line is added at the level of the amplitude at ipsilateral sites)

Fig. 5 Event-related lateralizations (ERLs) recorded in experiment 2 (intentional cueing). ERLs at a posterior electrode pair (PO7/ PO8) are depicted in the left and the middle panel for S–R corresponding (bold lines) and S–R non-corresponding (thin lines) trials. The peak in asymmetry is equivalent to the N2pc. In the left panel, ERLs are depicted stimulus coded to highlight differences in onset-latencies: posterior asymmetries started earlier in conditions where the side of the prepared response corresponded to the location of the target stimulus (valid intentional cueing S–R corresponding and invalid intentional cueing S–R non-corresponding; the effects are emphasized by dashed circles). Posterior ERLs in the middle panel and the LRP (right panel) are depicted response coded. Neither at posterior (N2pc) nor at central sites (LRP) was an effect of the type of cueing observed for the amplitude around the maximum of the early physic asymmetry

ERLs did not start earlier in validly cued trials in general, but whenever the cue pointed towards the location of the upcoming stimulus (in validly cued trials with corresponding S–R dimensions and in invalidly cued trials with non-corresponding S–R dimensions). In these conditions, the ERL rose about 31.5 ms earlier than when the (intentional) cue pointed away from the target position (162.5 vs 194 ms; cue validity by S–R correspondence: F(1,13)=15.622, p=.002). Additionally, a marginal main effect of cue validity (172 vs 184.5 ms; F(1,13)=4.436, p=.055), but no effect of S–R correspondence (F(1,13)<1), indicated slightly earlier onsets of the ERLs with validly cued responses. These results on early posterior asymmetries show a deficit, especially for trials in which the cue predicted neither the correct response nor the location of the upcoming imperative signal. Neither N2pc amplitude (F(2,26)=1.31, MSE= 10064.60, e=.92, p>.2) nor the early LRP component (F(2,26)<1) changed with the cueing condition. Discussion Behavioral data replicate previous findings on intentional cueing of the Simon effect (Hommel, 1996; Proctor et al., 1992; Proctor & Wang, 1997; Verfaellie et al.,

27

1988). The effect increases with valid advance information about the required response and decreases with invalid cueing compared to a regular Simon task (without cues). In contrast to behavioral data, electrophysiological correlates of processing the target stimulus, its localization, as well as the motor activation evoked by the localization process appeared to be unaffected by the intentional cue. The LRP in the cue– target interval indicated ongoing response preparation after the presentation of an intentional cue. The largely unchanged ERPs evoked by the target stimulus might be interpreted as indicators for a permanence of the mechanisms underlying the Simon effect even with intentional cueing. At first sight, this interpretation contradicts the increase in the effect in behavioral data. However, there is a change in the latency of the N2pc that might help to understand this contradiction. N2pc rose earlier whenever the side of the prepared response and the location of the upcoming target corresponded. This temporal aspect of the cortical asymmetries indicates perceptual acceleration with response–stimulus correspondence that might affect manual performance in the following way: validly cued corresponding trials as well as invalidly cued non-corresponding trials should be accelerated. Consequently, the Simon effect with valid cues should be enlarged and with invalid cues reduced, just as behavioral data has evidenced. Such a perceptual explanation for the influence of intentional cues upon the Simon effect was already proposed by Hasbroucq & Possama, (1994). They concluded that this finding indicates that the Simon effect might be due to processes involved in stimulus identification. However, in the present study, EEG measures indicate that the perception-based influence on the Simon effect with intentional cueing is not related to the visuomotor mechanisms that may basically underlie the Simon effect. These mechanisms are most probably reflected in the activation of motor areas evoked by the target stimulus (Praamstra & Plat, 2001), which is also unaffected by intentional cueing. According to the response selection account, we argued that the Simon effect should decrease with valid intentional cueing if the subjects selected the indicated response after the presentation of the cue. Only if the responses indicated by the cue are not selected but only a spatial code is activated, intentional cueing should not necessarily reduce the Simon effect. In the latter case, an increase of the effect with valid intentional cueing might also be in accordance with the response selection account. The LRP in the cue–target interval showed an early activation-related component that was similar to the asymmetry evoked in experiment 1 with attentional cueing. However, in contrast to experiment 1, the lateralized activity over the motor cortex sustained until the target was presented, indicating ongoing response preparation. In conclusion, ERP data showed that subjects did most probably select the response after the presentation

of an intentional cue5. Therefore, the fact that the Simon effect did not decrease with valid precues argues against response selection as the locus of interference between stimulus and response codes (see also Buckolz et al., 1996). Moreover, the increase in the Simon effect with intentional cueing might not be—even if the response was not selected in advance—an argument towards the response selection account, because the increase in the effect is most probably the consequence of perceptual factors, not directly related to the mechanisms evoking the Simon effect.

General discussion The present study tested the influence of attentional as well as intentional cueing upon the Simon effect. In previous studies, these variations of the Simon effect were conducted to obtain information first about the way the task-irrelevant spatial code is generated and second about the locus of interference causing the behavioral effect. According to the attentional-shift hypothesis (Nicoletti & Umilta`, 1994; Umilta` & Nicoletti, 1992), the Simon effect should disappear after cueing the most probable stimulus location because the shift of attention can occur already after the presentation of the informative cue. The subsequently presented target stimulus should evoke no more attentional shift and as a consequence no more Simon effect. The second assumption concerns the stage of interference. Most theories (e.g., Lu & Proctor, 1995; Proctor et al., 1992; Umilta` & Nicoletti, 1990) assume that the interference between irrelevant stimulus information and response information takes place within a response selection stage. According to this assumption, the Simon effect should disappear whenever a response is selected in advance, as is possible with valid intentional precues (Buckolz et al., 1996). Both predictions were, in accordance to previous studies, not supported by the data. Consequently, either the subjects did not perform properly with respect to the information delivered by the cues or the task as a whole has changed by using informative cues. Both notions can be excluded by electrophysiological measures. ERPs indicate that the subjects processed the information provided by the cue properly. With attentional cueing, changes in the amplitude and the latency of the target-evoked N2pc indicate that subjects had 5 One might argue that unless subjects are going to respond incorrectly on most invalid trials they have not selected the response based on the intentional cue. There is still a decision to be made, namely to execute the preselected response or to revise it (R. Proctor, personal communication, October, 2002). The crucial point here is how to define response selection. On the one hand, response selection might be defined as the final decision to execute a particular action. In this case, all responses, even in simple response tasks, have to be considered to require response selection. On the other hand, response selection may be defined as the generation or the retrieval of an S–R rule. In this case, subjects may use the selected response in validly cued trials and revise their decision in invalidly cued trials. We refer to the latter definition.

28

shifted their attention towards the cued location. With intentional cueing, sustained LRP activity in the cue– target interval—as a correlate for ongoing response preparation—was observed. The LRP evoked by the location of the target stimulus that was proposed as a valid indicator for behaviorally relevant response activation (Praamstra & Plat, 2001) remained unchanged across all tasks. Thus, regarding the generation of the irrelevant spatial code, the present data are in agreement with a referential coding (Hommel, 1993) rather than with an attentional-shift account. A constant frame of reference that is not affected by directing of attention might explain the persistence of the effect with attentional cueing. However, in the introduction, a number of fundamental concerns with the referential coding account were formulated, as, for instance, the referential coding approach cannot explain changes in the Simon effect evoked by modifications of the stimulus material that are known to be related to attention (e.g., the decrease in the effect with the presentation of a fixation cross; Proctor & Lu, 1994). Thus, neither the referential coding nor the attentional-shift account in their current forms provide a sufficient theoretical framework to explain the generation of the spatial code that is responsible for the Simon effect. A recent study (Ivanoff & Peters, 2000) proposed integrating these two accounts. They argue that a spatial code might be formed with respect to a shift of attention from an intentionally defined stimulus, i.e., that an object-based reference system as proposed by Hommel (1993) might be the origin of an attentional shift that is necessary but not sufficient for the Simon effect (lvanoff & Peters, 2000). Alternatively, the shortcomings of the attentionalshift account in explaining the persistence of the Simon effect with attentional cueing might be due to the attention model used (see Mu¨sseler, 1994). The attentional-shift account bases on the spotlight metaphor of visual attention (e.g., Posner, 1980), assuming that a spatially delimited spot moves around the visual field, from position to position. Alternatively, attention might be distributed across several positions in space simultaneously (e.g., Kramer & Hahn, 1995; LaBerge, Carlson, Williams, & Bunney, 1997). Attention within LaBerge’s model (LaBerge et al., 1997) is defined as consisting of two separate aspects: first a preparatory aspect that distributes weights of attention along a wide area of probable spatial location and second, the selective aspect that concentrates activity in a typical narrow area so that information within this area can be processed free of confusing information. When a stimulus falls into areas of high preparatory activity, selective activity is directed more rapidly to it than when the stimulus falls somewhere where activity is lower. Applying this theory to attentional cueing in the Simon task, one might assume that the cue might affect the preparatory aspect of attention only. The fact that no effects on the N1 component of the ERP were observed supports the notion that attention had not been

narrowed to the most probable target location at the moment the target appeared. Subsequently, the N2pc (Luck & Hillyard, 1994b) might reflect the narrowing of attention at the location of the relevant target stimulus. According to this idea, the onset of the N2pc was accelerated with valid attentional cueing and with intentional cueing if a response was prepared at the side where the target stimulus appears. The amplitude of the N2pc was increased after invalid attentional cues, indicating the increased effort to reallocate attention in this condition. Even though the attentional cue evoked a change in the attentional distribution across the visual field, a subsequent attentional shift, narrowing the focus at the target location might take place. Similar to the integration of the attentional-shift and referential coding account (Ivanoff & Peters, 2000) outlined above, processing of spatial information at two subsequent stages that requires a final narrowing of attention towards a lateralized location might be the crucial factor to obtain a Simon effect. Therefore, these data might still be in accordance with a revised attentional-shift account. The spatial code generated by this mechanism affects subsequent processing stages and thereby constitutes the Simon effect. The increase in the Simon effect with intentional cueing was initially interpreted to support the notion that interference takes place at a response-selection stage (Proctor & Wang, 1997). The present data question this interpretation. First of all, the Simon effect did not decrease even though subjects prepared a (necessarily) selected response in the cue target interval. Second, the increase in the Simon effect with intentional cues turned out to be most probably of perceptual origin, because ERP data showed that the intentional cue also affected stimulus processing at the cued location. The acceleration of stimulus processing in validly cued corresponding trials and invalidly cued non-corresponding trials can explain the increase in the Simon effect with valid intentional cueing without arguing towards a response-selection account of the Simon effect. Therefore, the present data add to previous evidence against the hypothesis that a response selection has to be made on the basis of the information delivered by the target stimulus to obtain a Simon effect. Hommel, (1996), for instance, also reported a Simon effect for simple responses whenever a simple response was followed by an alternate response. It was not the response selection itself, but holding a response alternative in evidence that appeared to be the crucial factor in obtaining a Simon effect. Thus, the interference between contradicting spatial information might take place after a response-selection stage as long as response selection is defined as the generation or retrieval of S–R mappings (see also footnote 5). Irrelevant spatial information could activate corresponding responses on a subsequent response-related stage, which might be responsible for the release of the final selected response or even for an execution stage. The irrelevant spatial code might bias response decision or execution rather than interfere within a response selection stage that provides S–R mappings. Only the

29

former mechanism should be unaffected by advance response selection, as indicated by the data presented. Acknowledgements We would like to thank Carlo Umilta`, Robert Proctor, and an anonymous reviewer for their constructive comments on earlier drafts of this paper. This study was supported by grants from the Deutsche Forschungsgemeinschaft to Edmund Wascher (Wa 987/5–1).

References Bisiacchi, P., Marzi, C.A., Nicoletti, R., Carena, G., Mucignat, C., & Tomaiuolo, F. (1994). Left-right asymmetry of callosal transfer in normal human subjects. Behavioural Brain Research, 64, 173–178. Buckolz, E., Odonnell, C., & McAuliffe, J. (1996). The Simon effect: evidence of a response processing ‘‘functional locus’’. Human Movement Science, 15, 543–564. Coles, M.G.H. (1989). Modern mind-brain reading: psychophysiology, physiology, and cognition. Psychophysiology, 26, 251–269. De Jong, R., Wierda, M., Mulder, G., & Mulder, L.J.M. (1988). Use of partial stimulus information in response processing. Journal of Experimental Psychology: Human Perception and Performance, 14, 682–692. De Jong, R., Liang, C.-C., & Lauber, E. (1994). Conditional and unconditional automaticity: a dual-process model of effects of spatial stimulus-response correspondence. Journal of Experimental Psychology: Human Perception and Performance, 20, 731–750. Eason, R.G. (1981). Visual evoked potential correlates of early neural filtering during selective attention. Bulletin of the Psychonomic Society, 18, 203–206. Eimer, M. (1994). ‘‘Sensory gating’’ as a mechanism for visuospatial orienting: electrophysiological evidence from trial-bytrial cuing experiments. Perception & Psychophysics, 55, 667–675. Eimer, M. (1995). Stimulus-response compatibility and automatic response activation—evidence from psychophysiological studies. Journal of Experimental Psychology: Human Perception and Performance, 21, 837–854. Eimer, M. (1996). The N2pc component as an indicator of attentional selectivity. Electroencephalography and Clinical Neurophysiology, 99, 225–234. Fitts, P.M., & Seeger, C.M. (1953). S-R compatibility: Spatial characteristics of stimulus and response codes. Journal of Experimental Psychology, 46, 199–210. Gratton, G., Coles, M.G.H., Sirevaag, E.J., Eriksen, C.W., & Donchin, E. (1988). Pre- and poststimulus activation of response channels: a psychophysiological analysis. Journal of Experimental Psychology: Human Perception and Performance, 14, 331–344. Hasbroucq, T., & Guiard, Y. (1991). Stimulus-response compatibility and the Simon effect: towards a conceptual clarification. Journal of Experimental Psychology: Human Perception and Performance, 17, 246–266. Hasbroucq, T., & Possamaı¨ , C.A. (1994). What can a precue enhance—an analysis of the experiments of Proctor, Lu and van Zandt (1992). Acta Psychologica, 85, 235–244. Hillyard, S.A., & Mu¨nte, T.F. (1984). Selective attention to color and location: an analysis with event-related brain potentials. Perception & Psychophysics, 36, 185–198. Hommel, B. (1993). The relationship between stimulus processing and response selection in the Simon task: evidence for a temporal overlap. Psychological Research, 55, 280–290. Hommel, B. (1996). S-R compatibility effects without response uncertainty. Quarterly Journal of Experimental Psychology Section A: Human Experimental Psychology, 49, 546–571. Ivanoff, J., & Peters, M. (2000). A shift of attention may be necessary, but it is not sufficient, for the generation of the Simon effect. Psychological Research, 64, 117–135.

Kornblum, S., Hasbroucq, T., & Osman, A. (1990). Dimensional overlap: cognitive basis for stimulus-response compatibility—a model and taxonomy. Psychological Review, 97, 253–270. Kornhuber, H.H., & Deecke, L. (1965). Hirnpotentiala¨nderungen bei Willku¨rbewegungen und passiven Bewegungen des Menschen: Bereitschaftspotential und reafferente Potentiale. Pflu¨gers Archiv fu¨r die gesamte Physiologie des Menschen und der Tiere, 284, 1–17. Kramer, A.F., & Hahn, S. (1995). Splitting the beam: distributions of attention over noncontiguous regions of the visual field. Psychological Science, 6, 381–386. Kutas, M., & Donchin, E. (1980). Preparation to respond as manifested by movement-related brain potentials. Brain Research, 202, 95–115. LaBerge, D., Carlson, R.L., Williams, J.K., & Bunney, B.G. (1997). Shifting attention in visual space: test of moving-spotlight models versus an activity-distribution model. Journal of Experimental Psychology: Human Perception and Performance, 23, 1380–1392. Lu, C.-H., & Proctor, R.W. (1995). The influence of irrelevant location information on performance: a review of the Simon and spatial Stroop effects. Psychonomic Bulletin & Review, 2, 174–207. Luck, S.J. (1995). Multiple mechanisms of visual-spatial attention: recent evidence from human electrophysiology. Behavioural Brain Research, 71, 113–123. Luck, S.J., & Hillyard, S.A. (1994a). Electrophysiological correlates of feature analysis during visual search. Psychophysiology, 31, 291–308. Luck, S.J., & Hillyard, S.A. (1994b). Spatial filtering during visual search: Evidence from human electrophysiology. Journal of Experimental Psychology: Human Perception and Performance, 20, 1000–1014. Luck, S.J., Hillyard, S.A., Mouloua, M., Woldorff, M.G., Clark, V.P., & Hawkins, H.L. (1994). Effects of spatial cuing on luminance detectability: psychophysical and electrophysiological evidence for early selection. Journal of Experimental Psychology: Human Perception & Performance, 20, 887–904. Mangun, G.R., & Hillyard, S.A. (1987). The spatial allocation of visual attention as indexed by event-related brain potentials. Human Factors, 29, 195–211. Marzi, C.A., Bisiacchi, P., & Nicoletti, R. (1991). Is interhemispheric transfer of visuomotor information asymmetric? Evidence from a meta-analysis. Neuropsychologia, 29, 1163–1177. Miller, J., Patterson, T., & Ulrich, R. (1997). A jackknife-based method for measuring LRP onset latency differences. Psychophysiology, 35, 99–115. Mu¨sseler, J. (1994). Position-dependent and position-independent attention shifts: evidence against the spotlight and premotor assumption of visual focussing. Psychological Research, 56, 251–260. Nicoletti, R., & Umilta`, C. (1994). Attention shifts produce spatial stimulus codes. Psychological Research, 56, 144–150. Oostenveld, R., Praamstra, P., Stegeman, D.F., & van Oosterom, A. (2001). Overlap of attention and movement-related activity in lateralized event-related brain potentials. Clinical Neurophysiology, 112, 477–484. Posner, M.I. (1980). Orienting of attention. Quarterly Journal of Experimental Psychology, 32, 3–25. Praamstra, P., & Plat, F.M. (2001). Failed suppression of direct visuomotor activation in Parkinson’s disease. Journal of Cognitive Neuroscience, 13, 31–43. Proctor, R.W., & Lu, C.-H. (1994). Referential coding and attention-shifting accounts for the Simon effect. Psychological Research, 56, 185–195. Proctor, R.W., & Wang, H. (1997). Enhancement of the Simon effect by response-location precues: evaluation of the stimulusidentification account. Acta Psychologica, 95, 279–298. Proctor, R.W., Lu, C.-H., & Van Zandt, T. (1992). Enhancement of the Simon effect by response precuing. Acta Psychologica, 81, 53–74. Rubichi, S., Nicoletti, R., Iani, C., & Umilta`, C. (1997). The Simon effect occurs relative to the direction of an attentional shift.

30 Journal of Experimental Psychology: Human Perception and Performance, 23, 1353–1364. Simon, J.R., & Rudell, A.P. (1967). Auditory S-R compatibility: the effect of an irrelevant cue on information processing. Journal of Applied Psychology, 51, 300–304. Steinmetz, H., Fu¨rst, G., & Meyer, B.-U. (1989). Craniocerebral topography within the international 10–20 system. Electroencephalography and Clinical Neurophysiology, 72, 499–506. Stoffer, T.H. (1991). Attentional focussing and spatial stimulusresponse compatibility. Psychological Research, 53, 127–135. Stoffer, T.H., & Umilta`, C. (1997). Spatial coding with reference to the focus of attention in S-R compatibility and the Simon effect. In B. Hommel & W. Prinz (Eds.), Theoretical issues in S-R compatibility (pp. 181–208). Amsterdam: North-Holland. Stoffer, T.H., & Yakin, A.R. (1994). The functional role of attention for spatial coding in the Simon effect. Psychological Research-Psychologische Forschung, 56, 151–162. Ulrich, R., & Miller, J. (2001). Using the jackknife-based scoring method for measuring LRP onset effects in factorial designs. Psychophysiology, 38, 816–827. Umilta`, C., & Liotti, M. (1987). Egocentric and relative spatial codes in S-R compatibility. Psychological Research-Psychologische Forschung, 49, 81–90. Umilta`, C., & Nicoletti, R. (1990). Spatial stimulus-response compatibility. In R.W. Proctor & T.G. Reeve (Eds.), Stimulusresponse compatibility (pp. 89–116). Amsterdam: Elsevier. Umilta`, C., & Nicoletti, R. (1992). An integrated model of the Simon effect. In J. Alegria & D. Holender (Eds.), Analytic approaches to human cognition (pp. 331–350). Amsterdam: North-Holland. Valle-Incla´n, F. (1996). The locus of interference in the Simon effect: an ERP study. Biological Psychology, 43, 147–162. Van der Lubbe, R.H.J., & Woestenburg, J.C. (1999). The influence of peripheral precues on the tendency to react towards a lateral relevant stimulus with multiple-item arrays. Biological Psychology, 51, 1–21.

Van der Lubbe, R.H.J., Wauschkuhn, B., Wascher, E., Niehoff, T., Kompf, D., & Verleger, R. (2000). Lateralized EEG components with direction information for the preparation of saccades versus finger movements. Experimental Brain Research, 132, 163–178. Verfaellie, M.D., Bowers, D., & Heilman, K.M. (1988). Attentional factors in the occurrence of stimulus-response compatibility effects. Neuropsychologia, 26, 435–444. Verleger, R., Gasser, T., & Mo¨cks, J. (1982). Correction of EEG artifacts in event-related potentials of the EEG: aspects of reliability and validity. Psychophysiology, 19, 472–480. Verleger, R., Vollmer, C., Wauschkuhn, B., Van der Lubbe, R.H.J., & Wascher, E. (2000). Dimensional overlap between arrows as cueing stimuli and responses? Evidence from contraipsilateral differences in EEG potentials. Cognitive Brain Research, 10, 99–109. Vogel, E.K., & Luck, S.J. (2000). The visual N1 component as an index of a discrimination process. Psychophysiology, 37, 190– 203. Wascher, E., & Wauschkuhn, B. (1996). The interaction of stimulus- and response-related processes measured by event-related lateralizations of the EEG. Electroencephalography and Clinical Neurophysiology, 99, 149–162. Wascher, E., Schatz, U., Kuder, T., & Verleger, R. (2001). Validity and boundary conditions of automatic response activation in the Simon task. Journal of Experimental Psychology: Human Perception and Performance, 27, 731–751. Yantis, S., & Jonides, J. (1990). Abrupt visual onsets and selective attention: voluntary versus automatic allocation. Journal of Experimental Psychology: Human Perception and Performance, 16, 121–134. Zimba, L.D., & Brito, C.F. (1995). Attention precuing and Simon effects—a test of the attention coding account of the Simon effect. Psychological Research-Psychologische Forschung, 58, 102–118.