Same/different concept learning in the pigeon: the effect of negative instances and prior adaptation to transfer stimuli.

Researchers taught pigeons to tell same from different pictures. Some birds saw only matching pairs. Others also saw pairs that did not match. Then all birds saw brand-new pictures to test if the rule still worked.

The team used matching-to-sample and oddity tasks. They tracked how fast each bird learned and how well they transferred the rule to new stimuli.

Birds that practiced with non-matching pairs learned the concept faster. When new pictures appeared, these birds picked the correct choice more often.

Seeing both matching and non-matching examples helped the pigeons form a clear same/different rule that worked with any pictures.

Catania et al. (1972) saw the opposite result: pigeons trained with only three picture layouts later failed when a fourth layout showed up. The difference is simple. C et al. never gave non-matching trials, so the birds never learned what "different" looked like.

Cook (2002) built on the 1978 finding. They showed that keeping the response rule the same across many picture sets makes the concept even stronger. Together, the two papers say: show both matches and non-matches, and keep the rule consistent.

Cohen (1969) and Bottjer et al. (1979) already proved pigeons can generalize matching. The 1978 study adds a teaching trick—include negative examples—to make that generalization happen sooner.

When you teach concepts like "same," "different," or "not," mix in clear non-examples. If a client is learning to sort "red versus not red," give equal practice with red items and items that are clearly not red. The brain needs both sides of the rule to transfer the skill to new objects, pictures, or words.