L. M. Hopper, C. L. Egelkamp, K. A. Cronin, S. L. Jacobson, K. E. Wagner and S. R. Ross
Lincoln Park Zoo, Lester E. Fisher Center for the Study and Conservation of Apes, Chicago, IL 60614, USA |
| The ability of primates to remember sequences is well documented. Less understood is what sequence-learning errors reveal about primate memory. Five gorillas and seven macaques, socially housed at Lincoln Park Zoo (Chicago, IL), were tested on a touchscreen serial-learning task. Subjects had to select symbols in a pre-determined order. After learning a two-item list (A-B), a third symbol (C) was added. When presented with the 3-item list, 31.1% of the subjects’ first 30 trials were correct on average. Rate of successfully sequencing the symbols varied by subject (P<0.001) but not by species or trial (both P>0.05). Success rate (mean=58.9%) was positively associated with selecting symbol A as the first list-item (P<0.001). However, in these trials subjects only subsequently chose symbol B, rather than C, at chance (mean=53.1%), with no variation across subjects (P>0.05). This suggests a failure to encode the last item (B) in the previously-learned sequence (A-B). The responses of one gorilla, who learned 4-, 5-, 6-, and 7-item lists, revealed the same encoding error: when he correctly sequenced n-2 items, he chose penultimate symbol at chance (mean=44.5%), with no effect of sequence length (P>0.05). The primates’ failure to encode the last item in a sequence, as revealed when a novel symbol was added in the ultimate position, supports the ordinal model of serial learning, such that encoding strength decreases by list-item position. |
