Our eye movements shift the visual scene on our retinas. These shifts go largely unnoticed and do not prevent efficient interaction with objects surrounding us. It has been proposed that the visual system compensates for such shifts using a copy of the motor command (Sperry, 1950) to anticipate changes in the visual scene from the planned eye movement. Such an active mechanism could maintain an impression of space constancy and allow us to effectively interact with visual objects. However, we typically do not keep track of the whole visual scene (O'Regan et al., 1999; Rensink et al., 1997). Studies have proposed that such visual compensation could be restricted to salient or task-relevant objects, selected by spatial attention (Cavanagh et al., 2010; Rolfs and Szinte, 2016). At the behavioral level, this compensation could result in anticipatory deployment of spatial attention to the retinal location that a visual stimulus will occupy after the saccade (Jonikaitis and Theeuwes, 2013; Rolfs et al., 2011; Szinte et al., 2015; Szinte et al., 2016). Such anticipatory deployment could explain observations that attention is allocated at a spatial target location almost immediately after a saccade (Jonikaitis et al., 2013; Yao et al., 2016b).

At the neuronal level, these visual compensations have been described as a remapping of visual neuron receptive fields. Remapping triggers an anticipatory and, sometimes, pre-saccadic response of neurons in frontal eye fields (FEF), lateral intra-parietal area (LIP) and superior colliculus (SC) with receptive fields centered on the post-saccadic retinal location of the attended object (Duhamel et al., 1992; Sommer and Wurtz, 2006; Walker et al., 1995). Remapping can facilitate tracking of task-relevant objects across saccades and allow rapid comparison between pre- and post-saccadic visual inputs (Crapse and Sommer, 2012). However, this remapping hypothesis has been challenged with new data collected within the FEF (Chen et al., 2018; Zirnsak and Moore, 2014). These studies found that, before a saccade, neurons respond to stimuli presented near the saccade target rather than to stimuli presented at remapped locations of the recorded receptive field (RF). These results were later termed ‘convergent remapping’ towards the saccade target in dissociation of the ‘forward remapping’, which would be parallel to the saccade vector (Neupane et al., 2016a). They led to the proposal that convergent remapping could manifest behaviorally as a spatially unspecific spread of attention around the saccade target (Zirnsak and Moore, 2014). Remapping of spatial attention before saccades, as reported in behavioral studies, therefore could be reinterpreted as attentional spread between saccade target and remapped location (Jonikaitis et al., 2013; Rolfs et al., 2011; Szinte et al., 2015; Szinte et al., 2016). Such interpretation of the convergent remapping effects predicts that locations surrounding the saccade target by up to 10 degrees of visual angle (dva) would receive all attentional benefits before the eyes start to move. To date, there are no behavioral studies mapping pre-saccadic attention in sufficient detail to disambiguate whether attention converges towards the saccade target, or is remapped in parallel to the saccade target, as earlier behavioral work suggested.

We developed a protocol that allowed us to measure detailed maps of pre-saccadic attention, by measuring the orientation sensitivity at multiple locations while participants prepared a saccade (Figure 1). We observed that attention was allocated to the saccade target location and did not spread to the nearby positions, about 4.2 degrees of visual angle (dva apart. Next, we measured remapping of attention in the presence of a salient cue during a saccade task, manipulating the timing of the cue relative to the saccade. Our reasoning was that if remapping is an attentional process (Cavanagh et al., 2010; Rolfs and Szinte, 2016), it will take some time for the attention shift to occur (Ling and Carrasco, 2006; Müller and Rabbitt, 1989; Nakayama and Mackeben, 1989; Rolfs and Carrasco, 2012). Therefore, stimuli presented just before a saccade would not leave enough time for remapping to develop and to be observed before the saccade. On the other hand, stimuli presented early enough should be remapped before the saccade. Indeed, we found that when the cue appeared shortly before saccade onset, spatial attention was allocated at the cued location but not at its remapped location. In contrast, when the cue appeared sufficiently early before saccade onset, attentional resources that were initially drawn to the cued location were re-allocated to its remapped location (i.e. the retinal location it will occupy after the saccade).

Figure 1

Download asset Open asset

We determined spatially detailed maps of attention before a saccade under two different conditions: first, when participants made a visually guided saccade, and second, when a transient peripheral stimulus, a cue, was additionally presented during its preparation. We assessed spatial attention by asking participants to report the orientation of a briefly presented tilted discrimination target (clockwise or counterclockwise tilted Gabor), embedded in a display of vertical distractor streams (vertical Gabors, Figure 1A–B). To ensure that the discrimination task could be solved correctly only if participants attended at a particular location, we first completed a threshold task in which participants fixated at the center of the screen. This threshold task was used to estimate the tilt angle of a cued discrimination target presented at different eccentricities from the fixation. We observed that to achieve comparable discrimination at different eccentricities, the discrimination target had to be tilted by 4.42 ± 0.86 ° (mean ± SEM), if presented at the fixation target. This tilt gradually increased with eccentricity, finally reaching 14.10 ± 1.40 ° at eccentricities between ~15.3 and~16.2 dva (see Figure 1C). We used these threshold tilt values at their respective eccentricities in the main saccade task.

We first verified that presentation of the discrimination target during saccade preparation did not disrupt eye movements. Such a disruption, as measured by saccade latency or accuracy, would suggest that the stimuli used to measure attention instead captured attention. For this we first determined whether eccentricity of the discrimination target affected saccade latency. Saccade latency was longer when the visual streams overlapped with the fixation and saccade targets (217.56 ± 3.77 ms) compared with when they didn’t overlap (186.00 ± 2.61 ms, p < 0.0001). This indicates that such difference resulted from the saccade target and fixation being less visible if they overlapped with the visual streams. Therefore, we separated the trials based on whether the fixation and saccade targets overlapped with the visual streams or not. Discrimination target eccentricity did not affect saccade latency on trials in which the fixation and saccade targets overlapped with visual streams. We did not observe a main effect of the discrimination target eccentricity (see Materials and methods for the definition of eccentricity), either for trials in which the fixation and saccade target overlapped with the visual streams (repeated measures ANOVA for four eccentricity groups used, F_3,39 = 0.08, p = 0.9725, ${\eta }_p^2$= 0.03 %) or for trials in which they didn’t (for four eccentricity groups used, F_3,39 = 1.49, p = 0.2312, ${\eta }_p^2$= 0.83 %). Note that from a pilot study, we expected to find such saccade latency costs when the targets overlapped with the visual streams. To compensate for these expected effects, on trials in which the visual streams overlapped with the fixation and the saccade target, we presented the discrimination target 25 ms later than when there was no overlap. This procedure ensured homogenous timing of the discrimination target relative to the saccade onset irrespective of the tested position. This also ensured that any spatially unspecific increase of the discrimination threshold before a saccade (Campbell & Wurtz, 1978) could not explain differences between our experimental conditions and that discrimination performance truly reflected sensitivity gathered at the same instant relative to the saccade onset. Next, we evaluated whether discrimination target eccentricity (i.e. the absolute distance between the saccade target and the saccade landing point) affected saccade accuracy, as would be evident if the target captured attention. Again, we didn’t find a main effect of the discrimination target eccentricity on the saccade accuracy (between five eccentricity groups used, F_4,52 = 2.11, p = 0.0929, ${\eta }_p^2$= 1.38 %). Altogether, these results show that presentation of the discrimination target did not disrupt saccade preparation, demonstrating that the stimuli we used to measure the allocation of attention did not directly interfere with its deployment.

Next, we measured the pre-saccadic allocation of attention. In all conditions, we analyzed performance to the presentation of discrimination targets within the 150 ms preceding the saccade. This procedure ensured that discrimination performance reflected the modulation of spatial attention over space rather than visual acuity and ensured that the same discrimination targets were matched across all conditions and trials. Trials with and without a cue were analyzed separately. On trials without a cue (Figure 2A–C), we found increased visual sensitivity (0.88 ± 0.05, normalized d’ and SEM, respectively) at the saccade target location relative to the average of all other tested positions (0.42 ± 0.04, p < 0.0001, Figure 2B), suggesting that attention shifted towards the saccade targets during saccade preparation. Further, we tested the spatial specificity of this effect by comparing visual sensitivity at the saccade target location with the average visual sensitivity at the four positions surrounding it. Attention at the saccade target clearly did not spread to the surrounding positions (0.40 ± 0.04, p < 0.0001 , Figure 2C), with sensitivity benefits being constrained to the immediate vicinity of the saccade target. Also, we did not observe a deployment of spatial attention to the fixation target (0.41 ± 0.05) compared with the average across all other positions (p = 0.8952).

Figure 2

Download asset Open asset

Next, we analyzed the trials during which we presented an additional cue during saccade preparation, with the cue and discrimination target shown shortly after each other (50 ms) and on average 96.88 ± 0.96 ms (cue offset relative to saccade onset) before the saccade (Figure 2D–F). Here, in addition to the fixation and saccade targets, we were also interested in two further locations: the cue location and the retinal location the cue will occupy after the saccade, that is, the remapped location. Visual sensitivity was higher at the cue (0.78 ± 0.07, p = 0.0002), at the saccade target (0.74 ± 0.07, p < 0.0001) and at the fixation target (0.55 ± 0.05, p = 0.0060), when compared to the average of all the tested locations (0.44 ± 0.03). When compared to their closest surrounding positions (Figure 2F), we found spatially specific effects only at the cue (surround: 0.37 ± 0.03, p < 0.0001) and at the saccade target (surround: 0.40 ± 0.03, p < 0.0001), but not at the fixation target (surround: 0.49 ± 0.03, p = 0.1112). Importantly, when the cue was shown shortly before the saccade, the visual sensitivity at its remapped location (0.46 ± 0.07) was not significantly higher relative to the other tested positions (0.44 ± 0.03, p = 0.6886). Thus, we observed no evidence for attentional remapping of the cued location when the cue appeared shortly before saccade onset.

These results contrasted with those found in trials when the same cue was shown substantially before the discrimination target (200 ms) and on average 240.82 ± 1.42 ms before the saccade onset (Figure 2G–I). Under such conditions, when compared with the average across all positions (0.45 ± 0.08), we found higher visual sensitivity at the saccade target (0.78 ± 0.08, p = 0.0004), at the cue (0.71 ± 0.06, p < 0.0001) and, critically, at the remapped location of the cue (0.56 ± 0.07, p = 0.0072). Similar to other experimental conditions, the benefits observed at the saccade target (surround: 0.46 ± 0.03, p = 0.0010), at the cue (surround: 0.41 ± 0.04, p < 0.0001) and at its remapped location (surround: 0.45 ± 0.04, p = 0.0382) did not spread towards their respective adjacent positions.

Finally, we found that the increase in sensitivity observed at the remapped location of the cue was present for the condition in which the cue appeared substantially before the discrimination target and the saccade onset, but not if it appeared later. Such an effect was evident from comparison of normalized sensitivity obtained at the remapped position of the cue in the condition in which no cue was shown (0.38 ± 0.05) to conditions in which the cue appeared substantially before the discrimination target and the saccade onset (0.56 ± 0.07, p < 0.0038) or to conditions in which it appeared later (0.46 ± 0.07, p = 0.2816). These comparisons can be visualized by mapping subtraction of the normalized sensitivity obtained in the conditions in which we displayed a cue from those in which no cue was shown (see Figure 3 and Materials and methods). We normalized these differences, to present data with the same color scale as in the condition maps (Figure 2B–2E–2H). Combined, our results so far show that when the visual system is given enough time to process and attend a visual stimulus, such as the salient cue used in our task, spatial attention is remapped to the retinal location the stimulus will occupy after a saccade.

Figure 3

Download asset Open asset

If attention is remapped pre-saccadically, one should also expect to find spatially specific attentional effects at the fixation target, as the fixation target is the remapped location of the saccade target. Indeed, one study reported such foveal remapping of the saccade target (Rolfs et al., 2011). In the experiment above, we observed inconsistent evidence for spatial attention at the fixation target (see Figure 2C, F and I). This is likely to be because of the threshold procedure differences between this study and the one that reported spatial remapping of attention to fixation (Rolfs et al., 2011). Indeed, as we were principally interested in remapping of the salient peripheral cue in the above study, our threshold procedure measured spatial attention during fixation (yielding, importantly, no threshold difference between the cue and the remapped location of the cue). However, previous research has shown that preparation of saccades draws spatial attention away from other, non-saccade or non-salient locations (Jonikaitis and Theeuwes, 2013), resulting in an increase of acuity threshold in between the fixation and the saccade target locations (Campbell and Wurtz, 1978). In the experiment described above, we measured perceptual thresholds in a fixation task, rather than in a saccade task, so it is possible that the threshold for the fixation location underestimated that during saccadic preparation. If this was the case, potential remapping of the saccade location to the fixation location could have been masked by drawing attention away from the fixation location in the threshold task. Therefore, in the threshold task of a second experiment (foveal remapping task), we adjusted the tilt of the discrimination target using a saccade rather than a fixation task (see foveal remapping main task in Materials and methods). We observed that the discrimination target had to be tilted by 15.03 ± 2.34° if presented at the fixation target before a saccade (Figure 1D), a tilt up to four times bigger than that recorded during the fixation threshold procedure (Figure 1C). We used these values in a simplified version of the above experiment, without presentation of the cue, and as before, with a discrimination target randomly shown across trials at 32 possible positions while participants prepared a visually guided saccade. If anything, this procedure should be even more sensitive at detecting attention spread between saccade and fixation targets (Zirnsak and Moore, 2014).

Then in this second task, in which only the thresholding procedure was changed, we first analyzed whether the discrimination target affected saccade latency and accuracy. We found a main effect of the eccentricity of the discrimination target on saccade latency within trials in which the visual streams overlapped with the fixation and the saccade targets (F_3,21 = 8.89, p = 0.0005, ${\eta }_p^2$= 4.84 %), but not on trials when the visual streams did not overlap with the targets (F_3,21 = 3.01, p = 0.0532, ${\eta }_p^2$= 2.03 %). Further, discrimination target eccentricity did not affect saccade accuracy (F_4,28 = 1.60, p = 0.2007, ${\eta }_p^2$= 2.53 %). These results suggest that, similar to the first experiment, presentation of the discrimination target had limited influence on preparation of the saccade. Next, and in agreement with the results of the first experiment, we found a systematic pre-saccadic deployment of attention towards the saccade target (Figure 4A, 0.68 ± 0.11, p < 0.0001) when compared with the average over all the tested positions (0.36 ± 0.03). Critically, this benefit was accompanied by systematic deployment of attention at the fixation target (0.96 ± 0.02, p < 0.0001). Finally, these effects were spatially specific (Figure 4B), as shown by the significant differences observed when comparing sensitivity at the fixation target (surround: 0.47 ± 0.06, p < 0.0001) and the saccade target (surround: 0.33 ± 0.04, p < 0.0001) with their relative surrounds.

We constructed spatial attention maps by measuring orientation sensitivity while participants prepared a visually guided saccade. Our paradigm allowed us to measure whether attention spreads broadly around the saccade target or shifts towards spatially specific loci. We observed that attention consistently shifted to the saccade target location and, importantly, did not spread to other locations surrounding it. In our main manipulation, we presented a salient cue during saccade preparation. On these trials, we observed a second spatially specific locus of attention at the cued location. Importantly, on these cued trials, we also observed a third, distinct locus of attention. Although this third location was neither salient, nor task-relevant before the saccade, it corresponded to the retinotopic location the cue would occupy after the saccade. In other words, we observed attentional remapping of the cue location before the saccade onset. Critically, these effects were present only when the cue appeared long enough before the saccade onset. This indicates that remapping, like any other attentional process, requires some time. As we observed three separate foci of attention (at the saccade target, cue and remapped location), we did not find support for the hypothesis that attention spreads during saccade preparation around the saccade target (Zirnsak and Moore, 2014).

Our findings speak to the current debate on whether forward remapping exists and what role it plays in maintaining visual stability. As we report behavioral results, it is important to note that the link between our effects and neurophysiology is only theoretical. Behavioral experiments do not allow for direct conclusions as to which neural structures are involved in remapping nor on the validity of neurophysiology findings. However, our findings are relevant to the behavioural predictions provided in earlier neurophysiology work. Combined behavioral and neuronal recordings are necessary to eventually link proposals of neuronal activity before saccades and behavioral observations reported here and previously.

Behavioral studies on visual remapping of attention (Jonikaitis et al., 2013; Rolfs et al., 2011; Szinte et al., 2015; Szinte et al., 2016; Yao et al., 2016b) have been inspired by earlier neurophysiology work suggesting that receptive fields in FEF, LIP and SC shift (or are remapped) in anticipation of post-saccadic stimulus location (Duhamel et al., 1992; Sommer and Wurtz, 2006; Walker et al., 1995). Such a predictive receptive field shift occurs either shortly after a saccade (at a neural latency too short for visual responses) or even before a saccade onset (Colby et al., 1996; Kusunoki and Goldberg, 2003). This so-called forward remapping offers an excellent candidate mechanism of space constancy that has been incorporated into such phenomenon models. In these models, connections between visual neurons with receptive fields spatially separated by the saccade vector can be used to predict post-saccadic visual stimulus location and compensate for retinal image shifts during the saccade (Neupane et al., 2017; Quaia et al., 1998; Wang et al., 2016). However, forward remapping models have been challenged by Zirnsak and colleagues (Chen et al., 2018; Zirnsak et al., 2014), whose results suggested that the early findings were affected by the low spatial resolution of the receptive field mapping technique (Duhamel et al., 1992; Sommer and Wurtz, 2006; Walker et al., 1995). More detailed FEF receptive field mapping suggested that, before a saccade, cells preferentially respond to stimuli presented near the saccade target rather than at the remapped target location. In other words, cell receptive fields do not shift in parallel to the saccade vector, but instead converge towards the saccade target. Zirnsak and colleagues thus argued that forward remapping models cannot explain space constancy and instead one should focus on saccade target selection as a mechanism mediating this phenomenon (Zirnsak and Moore, 2014). More recent work has indicated that the visual system may, in fact, implement both forward and convergent remapping of receptive fields in area V4 (Neupane et al., 2016b; Neupane et al., 2016a). As this combined approach has been criticized on technical grounds (Hartmann et al., 2017), the neurophysiological results regarding the existence of forward remapping remain inconclusive.

On the other hand, a number of behavioral studies have repeatedly and reliably demonstrated forward remapping of spatial attention before saccades (Jonikaitis et al., 2013; Rolfs et al., 2011; Szinte et al., 2015; Szinte et al., 2016; Yao et al., 2016b). The behavioral studies, however, suffered from the same drawbacks as early neurophysiological studies, low spatial resolution. Our study has eliminated this potential criticism, and our observed deployment of attention in the opposite direction of the saccade and of the cue speaks in favor of the forward remapping effects. Consistent with previous behavioral studies (Baldauf and Deubel, 2008; Deubel and Schneider, 1996; Jonikaitis et al., 2017) and contrary to the convergent remapping effects, our results show that spatial attention is allocated to the saccade target and does not broadly spread around it. Additionally, convergent remapping cannot account for a number of earlier behavioral findings (Jonikaitis et al., 2013; Rolfs et al., 2011; Szinte et al., 2015; Szinte et al., 2016), as such spread of attention would have to be asymmetric and not spread towards the several control positions tested in these earlier studies.

Of note, the behavioral consequences of pre-saccadic changes in the spatial tuning of visual cells RF are unclear. Perhaps counterintuitively, recent computational neuroimaging modeling (Kay et al., 2015) has shown that increasing the neural spatial sampling at a particular position, similar to over-sampling of the saccade target observed within convergent remapping studies (Hartmann et al., 2017; Tolias et al., 2001; Zirnsak et al., 2014), results in reduction of spatial uncertainty. Convergent remapping does not necessarily yield a large spread of attention around the saccade goal (Zirnsak and Moore, 2014). Rather, it may increase visual sensitivity to stimuli only in the immediate vicinity of the saccade target. Convergent remapping could, therefore, reflect the spatially specific attentional selection of the saccade target observed in the present study. Our results indicate that spatial visual attention mechanisms must be accounted for in future work of remapping to advance our understanding of space constancy. Here, we hypothesize that previous reports of convergent remapping may reflect increased visual sensitivity at the saccade target explained by both spatial and temporal properties of visual attention.

First, to determine a visual neuron RF spatial profile, neurophysiologists used localized visual probes, presented most of the time in a sparse display with high probe-background contrast. It is, therefore, likely that such probes capture spatial attention (Theeuwes, 1991). The same holds for visual stimuli used to trigger the saccade, which were both task-relevant and, in most experiments, high-contrast stimuli. As visual RFs can shift towards attended locations even without any saccade involved (Niebergall et al., 2011; Womelsdorf et al., 2006), one must account for the effect of attention before interpreting any RF change in spatial tuning. In our study, the attention-capturing cue and the saccade target were dissociated from the measure of spatial attention. We measured attention by a discrimination target that did not significantly capture attention on its own, and, therefore, did not interfere with saccade preparation or pre-saccadic distribution of spatial attention (Deubel and Schneider, 1996). We also use different conditions to account separately for the effect of the saccade target and of the cue. To understand pre-saccadic RF changes in spatial tuning, one should first consider the spatial deployment of spatial attention.

Second, we argue that the temporal dynamics of attention have to be accounted for. We found that the time at which we presented our cue was critical for remapping of attention to be observed before the saccade. In particular, benefits at the remapped location of the cue were observed only when the cue was shown more than 175 ms before the saccade onset. As even the fastest deployment of attention would take a minimum of 100 ms to occur (Ling and Carrasco, 2006; Müller and Rabbitt, 1989; Nakayama and Mackeben, 1989; Rolfs and Carrasco, 2012), our remapping effects are compatible with the time course of attentional selection. Different neurophysiology studies measured visual cell activity followingpresentation of visual objects at different times across saccades (Kusunoki and Goldberg, 2003; Marino and Mazer, 2018; Nakamura and Colby, 2002; Wang et al., 2016). In particular, it was reported that a fair proportion of visual neurons recorded within LIP (Kusunoki and Goldberg, 2003) and earlier visual areas (Nakamura and Colby, 2002) showed forward remapping activity for probes presented as early as 250 ms before the saccade. Interestingly, they observed forward remapping activity preceding the saccade onset only if a visual object was flashed early before the saccade, such that visual objects shown earlier relative to the eye movement resulted in delayed activity during or after the saccade onset (Kusunoki and Goldberg, 2003; Nakamura and Colby, 2002). Our results are consistent with these findings, but not with those of Wang and colleagues (Wang et al., 2016), which report that the presentation of a visual object before the saccade resulted only in post-saccadic forward remapping activities within a set of recorded cells in LIP. Moreover, they found that the recorded cells responded transiently to the intermediate position between the current and the future receptive field position. By contrast, we did not observe any benefit at locations between the cue and the remapped position of the cue. These effects suggest that remapping of attention better reflects the activity of early visual areas than attentional priority maps (LIP, FEF, SC).

It is important to note that neurophysiology studies that failed to observe pre-saccadic forward remapping typically presented probes shortly before (50 ms) the saccade. Our behavioral findings suggest that a short time window between probe presentation and neural recording, which some studies have used, might be insufficient for probes to be remapped to a location parallel to the saccade before the onset of the movement. Further, if remapping is closely related to the time course of attention, it is possible that for attended stimuli shown just prior to the saccade onset, remapping may occur during or after the saccade (Kusunoki and Goldberg, 2003; Nakamura and Colby, 2002). Indeed, Neupane and colleagues (Neupane et al., 2016a; Neupane et al., 2016b) observed forward remapping when measuring post-saccadic memory responses to probes shown just before the saccade. Also, Yao and colleagues (Yao et al., 2016a) have shown that a post-saccadic memory trace of remapping was enhanced by attentional modulations established before the saccade, corroborating the notion that forward remapping can occur after the saccade. Although we did not measure whether a cue shown shortly before saccade onset was remapped after the saccade in the present study, this was done in two previous studies (Jonikaitis et al., 2013; Yao et al., 2016b). Both studies found that spatial attention was allocated to the location of a salient stimulus immediately after the saccade, even when the stimulus was no longer present (Jonikaitis et al., 2013). This indicates that the visual system anticipates the attended stimulus location after a saccade and recomputes its retinotopic location before the saccade is done. Finally, we also observed high perceptual performance at fixation, a result in line with two previous studies that investigated foveal remapping effects (Knapen et al., 2016; Rolfs et al., 2011). Such an attentional effect is surprising, as one would expect that visual selection should prioritize the saccade target, whereas the current fixation should be the least informative and least attended part of the display (especially given that participants already fixated for ~1 s before starting saccade preparation). However, if one considers that fixation-centered receptive fields will process the saccade target after the saccade, forward remapping effects suggest significant attentional benefits at that location, as we observed here and in a previous study (Rolfs et al., 2011).

Marino and Mazer (Marino and Mazer, 2018) recently showed that attention modulates the state of neurons in V4 before saccade onset, without a substantial shift of neurons’ spatial tuning. In contrast to our study, which manipulated transient attention by cueing a location on every trial, they used a sustained attention task to a cue presented before a series of records. Their pre-saccadic effects are consistent with the results reported here and suggest that new models of space constancy should account for the dynamics of spatial attention.

In summary, we used an eccentricity-adjusted discrimination task to measure, for the first time, spatial attention maps before saccades. Using this method, we observed a spatially specific increase in visual sensitivity at the fixation target, the saccade target, the cue and the remapped location of the cue. We found no evidence supporting the convergent remapping interpretation, which suggests that spatial attention spreads around the saccade target in a spatially unspecific way. We found that, before a saccade, attention is deployed towards the saccade target as well as towards a cued location. Further, given that the cue was presented sufficiently early before the saccade, we observed a deployment of attention to its remapped location, that is parallel and opposite to the saccade vector. Although the benefit at that location is smaller compared with that at the cue location, it reflects an ongoing process that facilitates spatial attention allocation after the saccade despite the retinotopic shifts induced by the eye movement.

All files are available from the OSF database: URL: https://osf.io/3tru6.

1. 1. Szinte M
   2. Jonikaitis D
   3. Rangelov D

   (2018) Open Science Framework

   Dataset of the Peripheral Remapping and Foveal Remapping of attention tasks.

   https://osf.io/3tru6

1. 1. Baldauf D
   2. Deubel H

   (2008) Properties of attentional selection during the preparation of sequential saccades

   Experimental Brain Research 184:411–425.

   https://doi.org/10.1007/s00221-007-1114-x

   • PubMed
   • Google Scholar
2. 1. Brainard DH

   (1997) The Psychophysics Toolbox

   Spatial Vision 10:433–436.

   https://doi.org/10.1163/156856897X00357

   • PubMed
   • Google Scholar
3. 1. Campbell FW
   2. Wurtz RH

   (1978) Saccadic omission: why we do not see a grey-out during a saccadic eye movement

   Vision Research 18:1297–1303.

   https://doi.org/10.1016/0042-6989(78)90219-5

   • PubMed
   • Google Scholar
4. 1. Cavanagh P
   2. Hunt AR
   3. Afraz A
   4. Rolfs M

   (2010) Visual stability based on remapping of attention pointers

   Trends in Cognitive Sciences 14:147–153.

   https://doi.org/10.1016/j.tics.2010.01.007

   • PubMed
   • Google Scholar
5. 1. Chen X
   2. Zirnsak M
   3. Moore T

   (2018) Dissonant representations of visual space in prefrontal cortex during eye movements

   Cell Reports 22:2039–2052.

   https://doi.org/10.1016/j.celrep.2018.01.078

   • PubMed
   • Google Scholar
6. 1. Colby CL
   2. Duhamel JR
   3. Goldberg ME

   (1996) Visual, presaccadic, and cognitive activation of single neurons in monkey lateral intraparietal area

   Journal of Neurophysiology 76:2841–2852.

   https://doi.org/10.1152/jn.1996.76.5.2841

   • PubMed
   • Google Scholar
7. 1. Cornelissen FW
   2. Peters EM
   3. Palmer J

   (2002) The eyelink toolbox: eye tracking with MATLAB and the psychophysics toolbox

   Behavior Research Methods, Instruments, & Computers 34:613–617.

   https://doi.org/10.3758/BF03195489

   • Google Scholar
8. 1. Crapse TB
   2. Sommer MA

   (2012) Frontal eye field neurons assess visual stability across saccades

   Journal of Neuroscience 32:2835–2845.

   https://doi.org/10.1523/JNEUROSCI.1320-11.2012

   • PubMed
   • Google Scholar
9. 1. Deubel H
   2. Schneider WX

   (1996) Saccade target selection and object recognition: evidence for a common attentional mechanism

   Vision Research 36:1827–1837.

   https://doi.org/10.1016/0042-6989(95)00294-4

   • PubMed
   • Google Scholar
10. 1. Duhamel JR
    2. Colby CL
    3. Goldberg ME

    (1992) The updating of the representation of visual space in parietal cortex by intended eye movements

    Science 255:90–92.

    https://doi.org/10.1126/science.1553535

    • PubMed
    • Google Scholar
11. 1. Engbert R
    2. Mergenthaler K

    (2006) Microsaccades are triggered by low retinal image slip

    PNAS 103:7192–7197.

    https://doi.org/10.1073/pnas.0509557103

    • PubMed
    • Google Scholar
12. 1. Hartmann TS
    2. Zirnsak M
    3. Marquis M
    4. Hamker FH
    5. Moore T

    (2017) Two types of receptive field dynamics in area v4 at the time of eye movements?

    Frontiers in Systems Neuroscience 11:e13.

    https://doi.org/10.3389/fnsys.2017.00013

    • PubMed
    • Google Scholar
13. 1. Jonikaitis D
    2. Szinte M
    3. Rolfs M
    4. Cavanagh P

    (2013) Allocation of attention across saccades

    Journal of Neurophysiology 109:1425–1434.

    https://doi.org/10.1152/jn.00656.2012

    • PubMed
    • Google Scholar
14. 1. Jonikaitis D
    2. Theeuwes J

    (2013) Dissociating oculomotor contributions to spatial and feature-based selection

    Journal of Neurophysiology 110:1525–1534.

    https://doi.org/10.1152/jn.00275.2013

    • PubMed
    • Google Scholar
15. 1. Jonikaitis D
    2. Klapetek A
    3. Deubel H

    (2017) Spatial attention during saccade decisions

    Journal of Neurophysiology 118:149–160.

    https://doi.org/10.1152/jn.00665.2016

    • PubMed
    • Google Scholar
16. 1. Kay KN
    2. Weiner KS
    3. Grill-Spector K

    (2015) Attention reduces spatial uncertainty in human ventral temporal cortex

    Current Biology 25:595–600.

    https://doi.org/10.1016/j.cub.2014.12.050

    • PubMed
    • Google Scholar
17. 1. Knapen T
    2. Swisher JD
    3. Tong F
    4. Cavanagh P

    (2016) Oculomotor remapping of visual information to foveal retinotopic cortex

    Frontiers in Systems Neuroscience 10:54.

    https://doi.org/10.3389/fnsys.2016.00054

    • PubMed
    • Google Scholar
18. 1. Kusunoki M
    2. Goldberg ME

    (2003) The time course of perisaccadic receptive field shifts in the lateral intraparietal area of the monkey

    Journal of Neurophysiology 89:1519–1527.

    https://doi.org/10.1152/jn.00519.2002

    • PubMed
    • Google Scholar
19. 1. Ling S
    2. Carrasco M

    (2006) When sustained attention impairs perception

    Nature Neuroscience 9:1243–1245.

    https://doi.org/10.1038/nn1761

    • PubMed
    • Google Scholar
20. 1. Marino AC
    2. Mazer JA

    (2018) Saccades trigger predictive updating of attentional topography in area V4

    Neuron 98:429–438.

    https://doi.org/10.1016/j.neuron.2018.03.020

    • PubMed
    • Google Scholar
21. 1. Müller HJ
    2. Rabbitt PM

    (1989) Reflexive and voluntary orienting of visual attention: time course of activation and resistance to interruption

    Journal of Experimental Psychology: Human Perception and Performance 15:315–330.

    https://doi.org/10.1037/0096-1523.15.2.315

    • PubMed
    • Google Scholar
22. 1. Nakamura K
    2. Colby CL

    (2002) Updating of the visual representation in monkey striate and extrastriate cortex during saccades

    PNAS 99:4026–4031.

    https://doi.org/10.1073/pnas.052379899

    • PubMed
    • Google Scholar
23. 1. Nakayama K
    2. Mackeben M

    (1989) Sustained and transient components of focal visual attention

    Vision Research 29:1631–1647.

    https://doi.org/10.1016/0042-6989(89)90144-2

    • PubMed
    • Google Scholar
24. 1. Neupane S
    2. Guitton D
    3. Pack CC

    (2016a) Dissociation of forward and convergent remapping in primate visual cortex

    Current Biology 26:R491–R492.

    https://doi.org/10.1016/j.cub.2016.04.050

    • PubMed
    • Google Scholar
25. 1. Neupane S
    2. Guitton D
    3. Pack CC

    (2016b) Two distinct types of remapping in primate cortical area V4

    Nature Communications 7:e10402.

    https://doi.org/10.1038/ncomms10402

    • Google Scholar
26. 1. Neupane S
    2. Guitton D
    3. Pack CC

    (2017) Coherent alpha oscillations link current and future receptive fields during saccades

    PNAS 114:E5979–E5985.

    https://doi.org/10.1073/pnas.1701672114

    • PubMed
    • Google Scholar
27. 1. Niebergall R
    2. Khayat PS
    3. Treue S
    4. Martinez-Trujillo JC

    (2011) Expansion of MT neurons excitatory receptive fields during covert attentive tracking

    Journal of Neuroscience 31:15499–15510.

    https://doi.org/10.1523/JNEUROSCI.2822-11.2011

    • PubMed
    • Google Scholar
28. 1. O'Regan JK
    2. Rensink RA
    3. Clark JJ

    (1999) Change-blindness as a result of 'mudsplashes'

    Nature 398:34.

    https://doi.org/10.1038/17953

    • PubMed
    • Google Scholar
29. 1. Paradiso MA
    2. Carney T

    (1988) Orientation discrimination as a function of stimulus eccentricity and size: nasal/temporal retinal asymmetry

    Vision Research 28:867–874.

    https://doi.org/10.1016/0042-6989(88)90096-X

    • PubMed
    • Google Scholar
30. 1. Pelli DG

    (1997) The videotoolbox software for visual psychophysics: transforming numbers into movies

    Spatial Vision 10:437–442.

    https://doi.org/10.1163/156856897X00366

    • PubMed
    • Google Scholar
31. 1. Quaia C
    2. Optican LM
    3. Goldberg ME

    (1998) The maintenance of spatial accuracy by the perisaccadic remapping of visual receptive fields

    Neural Networks 11:1229–1240.

    https://doi.org/10.1016/S0893-6080(98)00069-0

    • PubMed
    • Google Scholar
32. 1. Rensink RA
    2. O'Regan JK
    3. Clark JJ

    (1997) To see or not to see: the need for attention to perceive changes in scenes

    Psychological Science 8:368–373.

    https://doi.org/10.1111/j.1467-9280.1997.tb00427.x

    • Google Scholar
33. 1. Rolfs M
    2. Jonikaitis D
    3. Deubel H
    4. Cavanagh P

    (2011) Predictive remapping of attention across eye movements

    Nature Neuroscience 14:252–256.

    https://doi.org/10.1038/nn.2711

    • PubMed
    • Google Scholar
34. 1. Rolfs M
    2. Carrasco M

    (2012) Rapid simultaneous enhancement of visual sensitivity and perceived contrast during saccade preparation

    Journal of Neuroscience 32:13744–13752.

    https://doi.org/10.1523/JNEUROSCI.2676-12.2012

    • PubMed
    • Google Scholar
35. 1. Rolfs M
    2. Szinte M

    (2016) Remapping attention pointers: linking physiology and behavior

    Trends in Cognitive Sciences 20:399–401.

    https://doi.org/10.1016/j.tics.2016.04.003

    • PubMed
    • Google Scholar
36. 1. Sommer MA
    2. Wurtz RH

    (2006) Influence of the thalamus on spatial visual processing in frontal cortex

    Nature 444:374–377.

    https://doi.org/10.1038/nature05279

    • PubMed
    • Google Scholar
37. 1. Sperry RW

    (1950) Neural basis of the spontaneous optokinetic response produced by visual inversion

    Journal of Comparative and Physiological Psychology 43:482–489.

    https://doi.org/10.1037/h0055479

    • PubMed
    • Google Scholar
38. 1. Szinte M
    2. Carrasco M
    3. Cavanagh P
    4. Rolfs M

    (2015) Attentional trade-offs maintain the tracking of moving objects across saccades

    Journal of Neurophysiology 113:2220–2231.

    https://doi.org/10.1152/jn.00966.2014

    • PubMed
    • Google Scholar
39. 1. Szinte M
    2. Jonikaitis D
    3. Rolfs M
    4. Cavanagh P
    5. Deubel H

    (2016) Presaccadic motion integration between current and future retinotopic locations of attended objects

    Journal of Neurophysiology 116:1592–1602.

    https://doi.org/10.1152/jn.00171.2016

    • PubMed
    • Google Scholar
40. 1. Theeuwes J

    (1991) Exogenous and endogenous control of attention: the effect of visual onsets and offsets

    Perception & Psychophysics 49:83–90.

    https://doi.org/10.3758/BF03211619

    • PubMed
    • Google Scholar
41. 1. Tolias AS
    2. Moore T
    3. Smirnakis SM
    4. Tehovnik EJ
    5. Siapas AG
    6. Schiller PH

    (2001) Eye movements modulate visual receptive fields of V4 neurons

    Neuron 29:757–767.

    https://doi.org/10.1016/S0896-6273(01)00250-1

    • PubMed
    • Google Scholar
42. 1. Walker MF
    2. Fitzgibbon EJ
    3. Goldberg ME

    (1995) Neurons in the monkey superior colliculus predict the visual result of impending saccadic eye movements

    Journal of Neurophysiology 73:1988–2003.

    https://doi.org/10.1152/jn.1995.73.5.1988

    • PubMed
    • Google Scholar
43. 1. Wang X
    2. Fung CC
    3. Guan S
    4. Wu S
    5. Goldberg ME
    6. Zhang M

    (2016) Perisaccadic receptive field expansion in the lateral intraparietal area

    Neuron 90:400–409.

    https://doi.org/10.1016/j.neuron.2016.02.035

    • PubMed
    • Google Scholar
44. 1. Womelsdorf T
    2. Anton-Erxleben K
    3. Pieper F
    4. Treue S

    (2006) Dynamic shifts of visual receptive fields in cortical area MT by spatial attention

    Nature Neuroscience 9:1156–1160.

    https://doi.org/10.1038/nn1748

    • PubMed
    • Google Scholar
45. 1. Yao T
    2. Treue S
    3. Krishna BS

    (2016a) An Attention-Sensitive memory trace in macaque MT following saccadic eye movements

    PLOS Biology 14:e1002390.

    https://doi.org/10.1371/journal.pbio.1002390

    • PubMed
    • Google Scholar
46. 1. Yao T
    2. Ketkar M
    3. Treue S
    4. Krishna BS

    (2016b) Visual attention is available at a task-relevant location rapidly after a saccade

    eLife 5:e18009.

    https://doi.org/10.7554/eLife.18009

    • PubMed
    • Google Scholar
47. 1. Zirnsak M
    2. Moore T

    (2014) Saccades and shifting receptive fields: anticipating consequences or selecting targets?

    Trends in Cognitive Sciences 18:621–628.

    https://doi.org/10.1016/j.tics.2014.10.002

    • PubMed
    • Google Scholar
48. 1. Zirnsak M
    2. Steinmetz NA
    3. Noudoost B
    4. Xu KZ
    5. Moore T

    (2014) Visual space is compressed in prefrontal cortex before eye movements

    Nature 507:504–507.

    https://doi.org/10.1038/nature13149

    • PubMed
    • Google Scholar

Author details

1. Martin Szinte

   Department of Cognitive Psychology, Vrije Universiteit, Amsterdam, The Netherlands

   Contribution

   Conceptualization, Resources, Data curation, Software, Formal analysis, Funding acquisition, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing

   For correspondence

   martin.szinte@gmail.com

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-2040-4005

2. Donatas Jonikaitis

   Department of Neurobiology, Howard Hughes Medical Institute, Stanford University School of Medicine, Stanford, United States

   Contribution

   Conceptualization, Investigation, Methodology, Writing—original draft, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-9851-0903

3. Dragan Rangelov

   Queensland Brain Institute, The University of Queensland, Brisbane, Australia

   Contribution

   Conceptualization, Resources, Investigation, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

4. Heiner Deubel

   Allgemeine und Experimentelle Psychologie, Ludwig-Maximilians-Universität München, Munich, Germany

   Contribution

   Conceptualization, Supervision, Methodology, Project administration, Writing—review and editing

   Competing interests

   No competing interests declared

• 1,963

  views

• 289

  downloads

• 34

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.