# Richard Lewis Elder, Ph.D.

The present report provides additional support for the existence of backward conditioning by eliminating long delay forward conditioning as an alternative explanation. Pigeons were autoshaped to peck at a vertical line under one of the following conditions: One group of subjects (backward conditioning group) had food presented on half the trials followed by a backward stimulus (red or green keylight); on the other half the trials, no food (lighted hopper only) was followed by the other color (green or red keylight). A second group of subjects (random control) received random presentations of the red or green stimulus following the food/no-food trials. For half the subjects in both these groups, the backward stimulus immediately followed the food or lighted hopper; for the other half of the subjects, there was a 10 second delay. An additional group (forward only control) never experienced a backward stimulus during training. After 8 days of autoshaping, all subjects were transferred to a discrete trial operant discrimination where the red and green keylights were the discriminative stimuli. For half the subjects in the backward conditioning groups (immediate and delay), the backward stimulus following food became the S$\sp{\rm D}$ and the backward stimulus that followed the lighted hopper (no food) became the S$\sp{\Delta}$. For the other half of the backward conditioning subjects, the backward stimulus following food became the S$\sp{\Delta}$ and the backward stimulus following the no food became the S$\sp{\rm D}$. For half the subjects in the random control and forward only control groups, the red was the S$\sp{\rm D}$ and the green was the S$\sp{\rm D}$, and the other half received the apposite combination. Dependent measures were the percent of responses to S$\sp{\rm D}$ and the difference between the responses to S$\sp{\rm D}$ and S$\sp{\Delta}$. With both dependent measures, the consistent immediate group learned the discrete trial operant discrimination faster than the other groups. The effect disappeared with the 10 second delay. The results are discussed as they relate to the criteria necessary for showing backward conditioning and theories that explain backward conditioning.