Research by Garcia and Koelling (1966) demonstrated that there are biological constraints to learning in classical conditioning. When testing taste aversion, they found that rats easily learned to associate a flavor with illness and subsequently avoided the flavor. However, the rats could not learn to associate lights and sounds with illness. This shows that organisms are biologically predisposed to associate certain stimuli (like flavor and illness) but not others, supporting the idea that conditioning helps organisms survive real health dangers.