Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Visual space is compressed in prefrontal cortex before eye movements



We experience the visual world through a series of saccadic eye movements, each one shifting our gaze to bring objects of interest to the fovea for further processing. Although such movements lead to frequent and substantial displacements of the retinal image, these displacements go unnoticed. It is widely assumed that a primary mechanism underlying this apparent stability is an anticipatory shifting of visual receptive fields (RFs) from their presaccadic to their postsaccadic locations before movement onset1. Evidence of this predictive ‘remapping’ of RFs has been particularly apparent within brain structures involved in gaze control2,3,4. However, critically absent among that evidence are detailed measurements of visual RFs before movement onset. Here we show that during saccade preparation, rather than remap, RFs of neurons in a prefrontal gaze control area massively converge towards the saccadic target. We mapped the visual RFs of prefrontal neurons during stable fixation and immediately before the onset of eye movements, using multi-electrode recordings in monkeys. Following movements from an initial fixation point to a target, RFs remained stationary in retinocentric space. However, in the period immediately before movement onset, RFs shifted by as much as 18 degrees of visual angle, and converged towards the target location. This convergence resulted in a threefold increase in the proportion of RFs responding to stimuli near the target region. In addition, like in human observers5,6, the population of prefrontal neurons grossly mislocalized presaccadic stimuli as being closer to the target. Our results show that RF shifts do not predict the retinal displacements due to saccades, but instead reflect the overriding perception of target space during eye movements.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Figure 1: Multi-electrode recordings and high-density mapping of FEF neuronal RFs.
Figure 2: Shifting of FEF neuronal RFs before saccade onset.
Figure 3: Fixation and presaccadic population RF densities of recorded FEF neurons.
Figure 4: Localization of probe stimuli by the population of recorded FEF neurons.


  1. Sommer, M. A. & Wurtz, R. H. Brain circuits for the internal monitoring of movements. Annu. Rev. Neurosci. 31, 317–338 (2008)

    Article  CAS  Google Scholar 

  2. Duhamel, J. R., Colby, C. L. & Goldberg, M. E. The updating of the representation of visual space in parietal cortex by intended eye movements. Science 255, 90–92 (1992)

    Article  ADS  CAS  Google Scholar 

  3. Walker, M. F., Fitzgibbon, E. J. & Goldberg, M. E. Neurons in the monkey superior colliculus predict the visual result of impending saccadic eye movements. J. Neurophysiol. 73, 1988–2003 (1995)

    Article  CAS  Google Scholar 

  4. Sommer, M. A. & Wurtz, R. H. Influence of the thalamus on spatial visual processing in frontal cortex. Nature 444, 374–377 (2006)

    Article  ADS  CAS  Google Scholar 

  5. Ross, J., Morrone, M. C. & Burr, D. C. Compression of visual space before saccades. Nature 386, 598–601 (1997)

    Article  ADS  CAS  Google Scholar 

  6. Kaiser, M. & Lappe, M. Perisaccadic mislocalization orthogonal to saccade direction. Neuron 41, 293–300 (2004)

    Article  CAS  Google Scholar 

  7. Robinson, D. A. & Fuchs, A. F. Eye movements evoked by stimulation of frontal eye fields. J. Neurophysiol. 32, 637–648 (1969)

    Article  CAS  Google Scholar 

  8. Moore, T. & Fallah, M. Control of eye movements and spatial attention. Proc. Natl Acad. Sci. USA 98, 1273–1276 (2001)

    Article  ADS  CAS  Google Scholar 

  9. Buschman, T. J. & Miller, E. K. Serial, covert shifts of attention during visual search are reflected by the frontal eye fields and correlated with population oscillations. Neuron 63, 386–396 (2009)

    Article  CAS  Google Scholar 

  10. Gregoriou, G. G., Gotts, S. J., Zhou, H. & Desimone, R. High-frequency, long-range coupling between prefrontal and visual cortex during attention. Science 324, 1207–1210 (2009)

    Article  ADS  CAS  Google Scholar 

  11. Gregoriou, G. G., Gotts, S. J. & Desimone, R. Cell-type-specific synchronization of neural activity in FEF with V4 during attention. Neuron 73, 581–594 (2012)

    Article  CAS  Google Scholar 

  12. Averbeck, B. B. & Lee, D. Coding and transmission of information by neural ensembles. Trends Neurosci. 27, 225–230 (2004)

    Article  CAS  Google Scholar 

  13. Ecker, A. S. et al. Decorrelated neuronal firing in cortical microcircuits. Science 327, 584–587 (2010)

    Article  ADS  CAS  Google Scholar 

  14. Dodge, R. The illusion of clear vision during eye movement. Psychol. Bull. 2, 193–199 (1905)

    Article  Google Scholar 

  15. Currie, C. B., McConkie, G. W., Carlson-Radvansky, L. A. & Irwin, D. E. The role of the saccade target object in the perception of a visually stable world. Percept. Psychophys. 62, 673–683 (2000)

    Article  CAS  Google Scholar 

  16. Deubel, H. & Schneider, W. X. Saccade target selection and object recognition: evidence for a common attentional mechanism. Vision Res. 36, 1827–1837 (1996)

    Article  CAS  Google Scholar 

  17. Rolfs, M. & Carrasco, M. Rapid simultaneous enhancement of visual sensitivity and perceived contrast during saccade preparation. J. Neurosci. 32, 13744–13752a (2012)

    Article  CAS  Google Scholar 

  18. Sheinberg, D. L. & Logothetis, N. K. Noticing familiar objects in real world scenes: the role of temporal cortical neurons in natural vision. J. Neurosci. 21, 1340–1350 (2001)

    Article  CAS  Google Scholar 

  19. Khayat, P. S., Spekreijse, H. & Roelfsema, P. R. Correlates of transsaccadic integration in the primary visual cortex of the monkey. Proc. Natl Acad. Sci. USA 101, 12712–12717 (2004)

    Article  ADS  CAS  Google Scholar 

  20. Khavat, P. S., Spekreijse, H. & Roelfsema, P. R. Visual information transfer across eye movements in the monkey. Vision Res. 44, 2901–2917 (2004)

    Article  Google Scholar 

  21. Bichot, N. P., Rossi, A. F. & Desimone, R. Parallel and serial neural mechanisms for visual search in macaque area V4. Science 308, 529–534 (2005)

    Article  ADS  CAS  Google Scholar 

  22. Moore, T. & Armstrong, K. M. Selective gating of visual signals by microstimulation of frontal cortex. Nature 421, 370–373 (2003)

    Article  ADS  CAS  Google Scholar 

  23. Ekstrom, L. B., Roelfsema, P. R., Arsenault, J. T., Bonmassar, G. & Vanduffel, W. Bottom-up dependent gating of frontal signals in early visual cortex. Science 321, 414–417 (2008)

    Article  ADS  CAS  Google Scholar 

  24. Noudoost, B. & Moore, T. Control of visual cortical signals by prefrontal dopamine. Nature 474, 372–375 (2011)

    Article  CAS  Google Scholar 

  25. Thompson, K. G., Bichot, N. P. & Sato, T. R. Frontal eye field activity before visual search errors reveals the integration of bottom-up and top-down salience. J. Neurophysiol. 93, 337–351 (2005)

    Article  Google Scholar 

  26. Connor, C. E., Preddie, D. C., Gallant, J. L. & Van Essen, D. C. Spatial attention effects in macaque area V4. J. Neurosci. 17, 3201–3214 (1997)

    Article  CAS  Google Scholar 

  27. Womelsdorf, T., Anton-Erxleben, K., Pieper, F. & Treue, S. Dynamic shift of visual receptive fields in cortical area MT by spatial attention. Nature Neurosci. 9, 1156–1160 (2006)

    Article  CAS  Google Scholar 

  28. Tolias, A. S. et al. Eye movements modulate visual receptive fields of V4 neurons. Neuron 29, 757–767 (2001)

    Article  CAS  Google Scholar 

  29. Nakamura, K. & Colby, C. L. Updating of the visual representation in monkey striate and extrastriate cortex during saccades. Proc. Natl Acad. Sci. USA 99, 4026–4031 (2002)

    Article  ADS  CAS  Google Scholar 

  30. Hamker, F. H., Zirnsak, M., Calow, D. & Lappe, M. The peri-saccadic perception of objects and space. PLOS Comput. Biol. 4, e31 (2008)

    Article  ADS  Google Scholar 

  31. Bruce, C. J. & Goldberg, M. E. Primate frontal eye fields. I. Single neurons discharging before saccades. J. Neurophysiol. 53, 603–635 (1985)

    Article  CAS  Google Scholar 

  32. Moore, T. & Fallah, M. Control of eye movements and spatial attention. Proc. Natl Acad. Sci. USA 98, 1273–1276 (2001)

    Article  ADS  CAS  Google Scholar 

  33. Hill, D. N., Mehta, S. B. & Kleinfeld, D. Quality metrics to accompany spike sorting of extracellular signals. J. Neurosci. 31, 8699–8705 (2011)

    Article  CAS  Google Scholar 

  34. Harris, K. D., Henze, D. A., Csicsvari, J., Hirase, H. & Buzsaki, G. Accuracy of tetrode spike separation as determined by simultaneous intracellular and extracellular measurements. J. Neurophysiol. 84, 401–414 (2000)

    Article  CAS  Google Scholar 

  35. Zirnsak, M., Lappe, M. & Hamker, F. H. The spatial distribution of receptive field changes in a model of peri-saccadic perception: predictive remapping and shifts towards the saccade target. Vision Res. 50, 1328–1337 (2010)

    Article  Google Scholar 

  36. Abbott, L. F. Decoding neuronal firing and modeling neural networks. Q. Rev. Biophys. 27, 291–331 (1994)

    Article  CAS  Google Scholar 

  37. Rao, J. S. & SenGupta, S. Topics in Circular Statistics. Series on Multivariate Analysis Vol. 5, Ch. 8, 175–204 (World Scientific Publishing, 2001)

    Google Scholar 

Download references


This work was supported by National Institutes of Health grant EY014924 and the Howard Hughes Medical Institute (T.M.). We thank D.S. Aldrich for technical assistance.

Author information

Authors and Affiliations



M.Z. and T.M. designed the study. M.Z., B.N. and K.Z.X. performed the experiments. M.Z. and N.A.S. analysed the data. M.Z., N.A.S. and T.M. wrote the manuscript.

Corresponding author

Correspondence to Marc Zirnsak.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Extended data figures and tables

Extended Data Figure 1 Schematic illustration of expected RF shifts due to ‘predictive remapping’.

a, Three RFs (highlighted circles A1, B1 and C1) during stable fixation at FP1. b, The same RFs as in a are shown during stable fixation at FP2 (A2, B2 and C2). In retinocentric areas like the FEF, RFs are displaced across fixations by the direction and amplitude equal to the saccade vector. c, According to predictive remapping, RFs are shifted from their presaccadic locations (grey circles) to their postsaccadic locations (gold circles) before movement onset. That is, RF shifts anticipate the upcoming eye movement.

Extended Data Figure 2 Saccadic reaction time.

Plotted are the average deviations from the mean saccadic reaction time of a given recording session for the two monkeys as a function of probe location.

Extended Data Figure 3 Time dependency of RF shifts.

a, Distribution of changes in shift amplitude (ΔPRE) between ‘early’ and ‘late’ PRFs (n = 68). Solid line indicates the mean (1.56 dva) of the distribution. b, Distribution of changes in the distance of PRFs (late–early) to the saccade target. Solid line indicates the mean (−1.26 dva) of the distribution.

Extended Data Figure 4 Retinocentric properties of FEF RFs (I).

a, Distribution of RF1s and RF2s in retinocentric coordinates shown together with the average retinocentric difference between corresponding centres across fixations. Error bars indicate s.d. b, Distribution of RF1s and RF2s projected into the same quadrant shown together with the average retinocentric difference between corresponding centres. This was done to control for possible systematic effects that might cancel each other out across quadrants in a. Error bars indicate s.d.

Extended Data Figure 5 Retinocentric properties of FEF RFs (II) and their presaccadic changes.

a, Correlations between the distances of RF1s to FP1 (εRF1) and the distances of RF2s to FP2 (εRF2) (r = 0.95, t-test (t) = 39.59, degrees of freedom (df) = 177, P < 10−10)* and between the angles ϕRF1 and θRF2 (r = 0.96, t = 45.32, df = 177, P < 10−10) are shown. Lines denote best regression fits. b, Correlations between εRF1 and the distances of PRFs to FP2 (εPRF) (r = 0.51, t = 7.81, df = 177, P = 2.1 × 10−8) and between the angles ϕRF1 and θPRF are shown. If the predictive remapping hypothesis were true, the relationships shown in a and b should not differ. However, the correlation between εRF1 and εPRF (b, top) is significantly different from the correlation between εRF1 and εRF2 (a, top) (Steiger’s Z-test for dependent correlations, Z = 13.4, P < 10−10), as are the intercepts b0 (t = 5.23, df = 283, P = 3.3 × 10−7) and the slopes b1 (t = 12.87, df = 283, P < 10−10) of the respective regressions. Furthermore, instead of a positive correlation close to 1 (a, bottom) we find a significant negative circular correlation** between ϕRF1 and θPRF (b, bottom) (r = −0.53, P < 10−10). c, Correlations between εRF1 and the distances of PRFs to FP1 (εPRF) (r = 0.43, t = 6.4, df = 177, P = 1.1 × 10−9) and between the angles ϕRF1 and ϕPRF (r = 0.68, t = 12.4, df = 177, P < 10−10) are shown. If there were no presaccadic shifts of RFs the relationships shown in a and c should not differ. Again, however, we find the correlation between εRF1 and εPRF (c, top) to be significantly different from the correlation between εRF1 and εRF2 (a, top) (Z = 14.03, P < 10−10), as are the intercepts (t = 20.81, df = 286, P < 10−10) and the slopes (t = 14.4, df = 286, P < 10−10) of the regressions. The same is true for the correlation between ϕRF1 and θRF2 (a, bottom) and the correlation between ϕRF1 and ϕPRF (c, bottom) (Z = 13.02, P < 10−10), and for the intercepts (t = 17.69, df = 352, P < 10−10) and slopes (t = 20.74, df = 352, P < 10−10). *Note, the significance of all correlations in this figure, reported as Pearson’s r, was also assessed by computing Spearman’s rho and Kendall’s tau. All correlations were significant using these measures (P values not reported). **Note, circular statistics (correlation and regression)37 that are independent of a particular coordinate system had to be used in this case as the plotted angles fell outside the linear region. For the relationship between ϕRF1 and θRF2 and between ϕRF1 and ϕPRF ordinary statistics could be used. For comparisons, the respective circular correlations are 0.96 and 0.71.

Extended Data Figure 6 Examples of RFs in which single probe results would be consistent with the remapping hypothesis.

Two examples of RF maps in which the presaccadic RF converges towards the saccade target, rather than remap, yet it is clear that sampling of responses from only a single location would yield results consistent with remapping. In each example, the two fixation and one presaccadic RF response maps, and corresponding RF centres (black crosses, RF1, RF2 and PRF), are shown from top to bottom, respectively. As in previous figures, the blue filled circle indicates the location of fixation during probe presentation, although in the presaccadic RF plot the monkey is preparing a saccade to FP2 (target). In addition, in each plot, the gold arrow denotes the vector describing the RF shift expected with remapping if that shift is exactly equal to the saccade vector (as in most studies). In addition, indicated along with the PRF of each example is the location to which the RF1 centre is expected to shift with remapping if the location is based on the empirically mapped postsaccadic RF (RF2) (white square). As FEF RFs are retinocentric, both predictions are virtually the same, but shown for clarity. Note that in both examples, although the PRF clearly deviates from the remapping prediction overall, the predicted remapping location nonetheless yields a clear visual response. Thus, if only a single probe is used, the results would be consistent with the remapping hypothesis. The vector plots below show the comparison of the empirical remapping prediction based on RF2 with the measured PRF shifts. Conventions are as in Fig. 1.

Extended Data Figure 7 Centres of all measured FEF RFs.

Distribution of RF1s (dark grey), RF2s (light grey) and PRFs (gold) are shown in head-centred (screen) coordinates. For RF1s and PRFs the monkey fixated FP1. For RF2s the monkey fixated FP2.

Extended Data Figure 8 Example of single neuron isolation, corresponding RF measurements, and waveform stability.

a, Density plot of isolated (left panel) and all other waveforms (right panel) recorded from a single U-Probe channel. b, Inter spike interval (ISI) histogram of the isolated waveform. The number of ISI violations and the estimated false positive rate are based on a refractory period of 1.5 ms. b, Averaged (arithmetic mean) waveforms (solid lines) for the fixation (left) and presaccadic (middle) condition, and their differences (right). Grey and gold dashed lines indicate s.d. Black dashed lines enclose 95% confidence intervals. None of the depicted differences are statistically significant. d, Projection of fixation (grey) and presaccadic (gold) waveforms into principal component (PC) space. First two PC dimensions are shown. e, Distribution of Pseudo R2s (n = 40) resulting from logistic regression fits for each PC dimension in order to separate between fixation and presaccadic waveforms. Dnull designates the null deviance using the intercept in the regression exclusively. Dk designates the model deviance using a single PC dimension as predictor in addition. Perfect seperation would result in an R2 of 1. The mean of the depicted R2s is 0.01 (min = 2.9 × 10−6, max = 0.04). None of the R2s reached statistical significance (likelihood ratio chi-squared test). f, Performance of a linear support vector machine trained to discriminate between fixation and presaccadic waveforms using all 40 PC dimensions simultaneously. The estimated performance (gold line) of 59.4% correct classifications falls well within the 95% range (37.1, 62.9) (red lines) of the expected chance performance. g, Fixation and presaccadic RF maps are shown together with their respective RF centres (black crosses, RF1, RF2 and PRF). Same conventions as in Fig. 1. h, Histogram of the pseudo R2s as obtained by logistic regressions for 21 neurons with stable waveforms. The mean R2 is 0.01 (min = 0, max = 0.23). None of the individual R2s reached statistical significance. i, Support vector machine classification performance of fixation and presaccadic waveforms for the same set of neurons. The mean correct classification performance is 50.1% (s.d. = 7.3). All performance estimates (gold ellipses) fall well within the expected ranges of chance performance (grey lines).

Extended Data Figure 9 RF shifts of single neurons and comparison with RF shifts in the remaining population.

a, The presaccadic shift amplitude (ΔPRE) as a function of the distance of RF1 from the saccade target (FP2) (εRF1) for well-isolated single neurons with stable waveforms (left) and for the remaining set of RFs (right). Lines denote fits of a linear regression. b, The angular deviation of the presaccadic RF shift from the remapping prediction (ϕ′) as a function of θ for the two subpopulations of RFs. c, Comparison of the population of presaccadic RF shifts (gold vectors) with the remapping prediction (grey vectors) for the two subpopulations of RFs. Vector origins and end points are based on, respectively, RF1 and RF2 (grey) and RF1 and PRF (gold).

Extended Data Figure 10 Spike count correlations of recorded FEF neurons.

a, Decrease in mean spike count correlation (rSC) as a function of electrode distance. Solid line denotes the best fitting power function (axb; a = 0.12, P < 10−10; b = −0.45, P < 10−10). Data points are averages across neuronal pairs (n = 677) recorded at a fixed electrode distance, across all probe locations and across all three experimental conditions (fixation 1, fixation 2 and presaccadic). The number of neuronal pairs for each electrode distance, in order of increasing distance, was n = (128, 106, 92, 86, 68, 55, 44, 28, 24, 18, 14, 10, 4). Note that the fit was based on the non-averaged data. b, Top, mean rSC plotted as a function of baseline normalized firing rate during fixation (grey) and before saccade onset (gold). Positive values on the abscissa indicate combined responses above baseline. Error bars indicate s.e.m. Number of combined responses, in order of increasing baseline normalized response, was n = (8,805, 8,805, 8,806, 8,804, 8,806, 7,738, 7,737, 7,738, 7,737, 7,737, 7,738, 7,737, 7,738, 7,737, 7,737) for fixation, and n = (2,827, 1,742, 1,416, 1,185, 1,144, 1,026, 1,298, 1,655, 2,511, 3,876, 6,013, 7,342, 7,679, 9,013, 11,329) for the presaccadic condition. Middle, cumulative percentage of paired neuronal responses for which each individual response exceeds baseline during fixation (grey) and before saccade onset (gold). Bottom, difference in the mean rSC during fixation and before saccades. Error bars indicate Bonferroni-corrected (15 comparisons) 95% confidence intervals. Grey area indicates nonsignificant differences.

PowerPoint slides

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Zirnsak, M., Steinmetz, N., Noudoost, B. et al. Visual space is compressed in prefrontal cortex before eye movements. Nature 507, 504–507 (2014).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

This article is cited by


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing