Abstract
The extent to which different species display self-recognition is a controversial topic in comparative cognition. Self-recognition is widely validated through the mark test in which a dye or paint mark is applied surreptitiously to the subject. Mark-directed responses in the presence of a mirror are taken as evidence of self-recognition. Over the past 50 years many different species have been administered the mark test, but only a handful have passed it. Some have suggested that species that pass the mark test are those that have a complex social structure. In this review, we propose, rather, that the reason why an animal fails the mark test is because the study has failed to adequately account for one or more of three crucial elements: mirror understanding, self-exploration, and mark saliency. Alternative methods of validating mirror self-recognition also are discussed.