Detecting metamemory in nonverbal subjects: A test with pigeons

We describe a novel, nonverbal test of metamemory. In a 3-alternative delayed matching-to-sample task, pigeons were occasionally allowed to choose between a test of memory for the sample with 6 pellets reinforcement for correct matching and pecking a safe key for 3 pellets. Birds increasingly chose the safe key over the test of memory as the retention interval increased. In addition, matching was more accurate when birds chose the test of memory on trials with the safe key than when they were forced to have their memory tested, as if they monitored memory strength on each trial. However, when the birds had to commit to a test of memory or the safe key before the sample and distracters appeared, the data were more consistent with an alternative interpretation. This type of experimental design can be used more generally to test for metacognitive processes in nonverbal species and has several advantages over designs reported previously.