Figure 1.
This figure shows schematic diagrams of the visual matching-to-sample tasks used in the current study.
The top panel depicts the task used in Experiment 1 and as baseline in Experiment 2 (Standard visual matching trials). The lower panel depicts test trials in Experiment 2 (Cross-modal probe trials).
Figure 2.
Line Graph: Generalization to novel videos in Experiment 1A.
The uppermost line (diamond symbols, dashed line) represents trials with all familiar videos from previous phases of training, only during test sessions in which new videos were introduced. The lower data line (square symbols, solid line) represents performance on the first exposure to newly introduced videos. The bottom dashed line indicates accuracy expected by chance. Bars: Generalization to test trials in Experiment 1A and Experiment 1B. The bar with checker pattern represents performance in probe trials with novel videos depicting stimulus monkeys from the novel perspective of the back of a housing cage, during a single test session (Experiment 1A). The hatched bar represents performance in probe trials in which the screen went black and choice images appeared immediately after the video ended.
Figure 3.
Proportion of correct choices on trials with a vocalization from the monkey seen in the sample video (Congruent), without a vocalization (Control), or with a vocalization from a monkey other than the one seen in the sample video (Incongruent) in Experiment 2.
Accuracy expected by chance is 0.2.
Figure 4.
Proportion of errors made by selecting the vocalizing monkey in incongruent trials from Experiment 2.
The dashed line represents the proportion of choices expected by chance.