De Waal argues that we should attempt to understand a species’ intelligence only within its own context, or umwelt: the animal’s “self-centered subjective world, which represents only a small tranche of all available worlds.” There are many different forms of intelligence; each should be valuated only relative to its environment. “It seems highly unfair to ask if a squirrel can count to 10 if counting is not really what a squirrel’s life is about,” de Waal writes. (A squirrel’s life is about remembering where it stored its nuts; its intelligence is geospatial intelligence.) And yet, there’s apparently a long history of scientists ignoring this truth. For example, they’ve investigated chimpanzees’ ability to recognize faces by testing whether the chimps can recognize human faces, instead of faces of other chimps. (They do the former poorly and the latter quite well.) They’ve performed the famous mirror test — to gauge whether an animal recognizes the figure in a mirror as itself — on elephants using a too-small, human-size mirror. Such blind spots are, ultimately, a failure of empathy — a failure to imagine the experiment, or the form of intelligence it’s testing for, through the animal’s eyes. De Waal compares it to “throwing both fish and cats into a swimming pool” and seeing who can swim.