When pros become cons for anti- versus prosaccades: factors with opposite or common effects on different saccade types

Abstract

In five experiments we compared prosaccade and antisaccade performance in normal human observers. This was first examined for visual stimulation in temporal or nasal hemifields, under monocular viewing. Prosaccades were faster following temporal than nasal stimulation, in accordance with previous results. The novel finding was that the opposite pattern was observed for antisaccades, consistent with a difficulty in overcoming a stronger prosaccade tendency after temporal-hemifield stimulation. A second experiment showed that these results were not simply due to antisaccades following nasal stimulation benefitting from being made towards a temporal place-holder. Prosaccades and antisaccades were then compared for visual versus somatosensory stimulation. The substantial latency difference between prosaccades and antisaccades for visual stimuli was eliminated for somatosensory stimuli. Antisaccades can thus benefit in relative terms when the competing prosaccadic tendency is weakened; but two further experiments revealed that not all manipulations induce opposing outcomes for the two types of saccade. Although reducing the contrast of visual targets can slow prosaccades and conversely speed antisaccades, this was not the case at the lowest contrast level used, where both types of saccade were slowed, thus indicating some common limiting source. Moreover, warning sounds presented shortly before a visual target speeded both prosaccades and antisaccades. These results illustrate that several factors which slow prosaccades can speed antisaccades (consistent with competition between different pathways); but also reveal some notable exceptions, where both types of saccade are slowed or speeded together, suggesting some common pathways that may precede competition over the direction of the saccade.

A competitive integration model of exogenous and endogenous eye movements

“…Figure 5 shows that, while prosaccades become faster when the target has a higher strength, antisaccades are slowed down. This is in line with findings from Kristjánsson et al (2004). These authors also found that for very low-contrast targets, both pro-and antisaccades became slower.…”

“…This simulation Kristjánsson et al (2004) showed that the lack of inhibition resulted in mostly erroneous prosaccades to the onset location. When no prosaccade was made, antisaccades tended to be very slow as the resolution of competition between the pro-and antisaccade goal took very long.…”

“…These prosaccades are thus mostly very fast saccades: 4.5% of saccades in the antisaccade paradigm were fast erroneous prosaccades (9.7% in the antisaccade paradigm with gap). In most antisaccade studies, erroneous prosaccades are made on 10-20% of trials, and indeed tend to be the fastest saccades (e.g., Bell et al 2000;Everling et al 1999;Kristjánsson et al 2004;Mokler and Fischer 1999). Observers usually also produce some prosaccades with a normal latency.…”

The effect of attentional demands on the antisaccade cost

“…In a follow-up study, Kristjánsson et al (2004) demonstrated that certain manipulations had opposite effects on prosaccades and antisaccades, which provided further evidence for the competition account of antisaccade generation.…”

“…However, several findings from Kristjánsson and colleagues (Kristjánsson, 2007;Kristjánsson et al, 2001;Kristjánsson et al, 2004) have provided evidence for the view that prosaccades and antisaccades are not programmed independently, but instead compete interactively. For example, Kristjánsson et al (2001) showed that when attending to a secondary task (a peripheral discrimination task) shortly before the presentation of the antisaccade stimulus, latencies were increased in the prosaccade task but reduced in the antisaccade task.…”