Cross-modal recognition by the monkey: can the experience in the first modality be reduced to a single trial?

Using the cross-modal recognition procedure previously described, but now progressively reducing the number of learning trials given in the first sense modality, three unselected and unoperated rhesus monkeys (and one with bilateral removals of frontal, temporal and parietal polysensory cortex) were able to maintain an average level of 75% correct cross-modal recognition with only 1-3 learning trials. The pool size seemed less important than the familiarity of the objects. No discontinuity between single-trial and multiple-trial learning procedures was detected. Instead, the number of objects available at one time during learning was regarded as critical for distinguishing between these two kinds of learning procedure.