Saccadic latencies are known to change as a function of target eccentricity. Recently, it has been shown that latencies may be evaluated in terms of the amplitude of the step in proportion to the size of the target, and consistently change according to this step-size ratio with smaller ratios producing longer latencies (Madelain et al., 2005; Harwood et al., 2008; De Vries et al., 2016). This effect, called the size-latency phenomenon, might be seen as a function of a cost-benefit relationship: the difference in latencies might be explained by the ‘cost’ of making a saccade while the target mostly remains within the attentional field.
Here, we probe this hypothesis by manipulating the cost-benefit relationship using a reinforcement procedure. Six subjects (including two authors) tracked a visual ring target stepping horizontally with an amplitude ranging from 1.2 to 10.5 deg. The size (diameter) of the ring varied as a function of the target step such that the step-size ratio was equal to either 0.3 or 1.5. Trials with saccadic latencies outside a [80;500] ms range or saccadic gains outside [0.5;2] were discarded.
We used a dynamic reinforcement criterion based on the median computed over a 50-trial moving window in 2 blocked conditions. In the 0.3 ratio condition, any latency shorter than the criterion was reinforced. In the 1.5 ratio condition, any latency longer than the criterion was reinforced.
During baseline, we observed the size-latency phenomenon with large differences in latencies depending on the ratio (e.g. 152 ms and 204 ms, respectively for 1.5 and 0.3). After training (4800 reinforcement trials), distributions shifted toward the shorter or longer value (e.g. 223 ms and 169 ms, respectively for 1.5 and 0.3). On average, latencies decreased by 31 ms and increased by 75 ms according to the ongoing reinforcement contingencies. These changes in saccadic latencies were not explained by changes in saccadic amplitudes.
Reinforcement reduced the size-latency phenomenon, although it was not entirely suppressed. Our results indicate that reinforcement contingencies can considerably affect saccadic latency distributions, and support the idea of a cost-benefit evaluation for saccade triggering.