Representation of serial order in humans: A comparison to the findings with monkeys ( <Emphasis Type="Italic">Cebus apella </Emphasis>)

The ability to acquire serially ordered information and use that information in a flexiblemanner has been the subject of considerable investigation in pigeons, monkeys, and humans, and it is important for our understanding of many other behaviors such as language, maze learning, timing,motor control, and memory. Of the many types of tasks that tap serially organized behavior, perhaps the one that has enjoyed the greatest success in revealing differences across species is the simultaneous-chaining serial-order paradigm. Originally developed by Terrace for use in pigeons (Straub, Seidenberg, Bever, & Terrace, 1979; Straub & Terrace, 1981) and later adapted for use in New World (Cebus apella) monkeys (D’Amato & Colombo,1988), the procedure is quite simple.At the start of a trial, five stimuli are presented simultaneously. For ease of exposition, they may be referred to as A, B, C, D, and E. In our studies (D’Amato & Colombo, 1988, 1989, 1990), the monkeys were trained to respond to the five stimuli in a specific order—namely, A®B®C®D®E. Although on each trial the stimuli appeared in different spatial positions, the order of correct responding(A®B® C®D®E) remained the same. Any deviation from the assigned response order, such as by skipping over an item that should have been pressed (e.g., A®B®D), or responding to an item that had already been pressed (e.g., A®B®C®A), was considered incorrect and terminated the trial. Once the monkeys had acquired the serial-order task to a level of 75% correct or better over 2 consecutivedays, they were given a pairwise test in which only two of the five stimuli were presented on each trial. With these five stimuli, 10 possible pairs could be generated (AB, AC, AD, AE, BC, BD, BE, CD, CE, and DE). We found that the monkeys responded to the stimuli in the pairs in the order in which they appeared in the original sequence. When presented with pair BD, for example, the monkeys f irst pressed stimulus B and then pressed stimulus D (D’Amato & Colombo, 1988). The high level of performance on this pairwise test suggested that in learning the original serial-order task, the monkeys had formed an internal linear representationof the series.When confronted with a test pair, the monkeys accessed the internal representation at A and then proceeded through the representation until an item in the representation matched an item on the display. If indeed the monkeys formed and accessed a representation in the manner described, then the latency to respond to the first stimulus of a test pair should increase as a function of whether the first stimulus pressed was A, B, C, or D. For example, the latency to respond to stimulus B in pair BD, which requires accessing only one item (A) prior to B, should be much shorter than the latency to respond to stimulusC in pair CD,which requires accessing two items (A and B) prior to C. This is exactly what we noted. Overall, there was a strong linear relationship between the latency to respond to the first item of a pair This research was supported by a University of Otago Division of Sciences Grant to M.C. Address correspondence to M. Colombo, Department ofPsychology,University ofOtago,Dunedin,New Zealand (e-mail: [email protected]). BRIEF REPORTS