From Diana McSherry, Houston, Texas, US
Continuing Conor Feehly’s interesting points in his article “What makes a mind?”, I would propose that “agency” isn’t an adequate definition of “self”. As Feehly points out, the concept of agency is complicated because comprehensive, inanimate systems can make complex decisions based on novel input. But we don’t attribute agency to neural network software programs. Nor does science attribute agency to evolution, which is driven by natural selection (14 February, p 32).
A test that established an entity had a sense of “self”, separating it from the environment, would be useful. The mirror test, where an animal recognises itself, would be a good example. But pets don’t normally pass this test, and my friends insist that pets not only have a well-established sense of self, but can also be incredibly manipulative. Continuing into the insect world, bees can lead other bees to a novel source of food, so should they be included in having a sense of self?
