Shared attention in pigeons.

Bennett et al. (1973) taught pigeons a matching game. Birds saw either one simple light or two lights that overlapped. They had to peck the key that matched the sample.

The team tested many exposure times to see if longer looks helped when the sample was two lights blended together.

Pigeons made more errors when the sample was a blended pair of lights. Single-light samples gave sharper stimulus control.

Even when birds stared longer at the compound, accuracy stayed lower. Shared attention hurt performance at every duration.

Catania (1973) ran the same task the same year and got the opposite result. C's birds did better with redundant compounds. The key difference: C's lights did not overlap; either color, line tilt, or both could predict the answer. Overlap seems to split attention instead of doubling it.

Thrailkill et al. (2025) later showed that reinforcement history with single elements biases later attention to any compound that contains them. This extends the 1973 warning: past rewards can further dilute control when compounds appear.

Snapper et al. (1969) had already shown pigeons can learn color-form compounds, but form cues usually win. S et al. now quantify the cost of that competition.

When you teach conditional discriminations, keep target stimuli separate at first. Blended or overlapping visuals split the learner's attention and weaken stimulus control. Start with clean single cues, then combine only after each element holds strong control. Check for overlap in your pictures, icons, or video prompts—less overlap, better matching.