Counting and indeterminate identity

Suppose that we repair a wooden ship by replacing its planks one by one with new ones while at the same time reconstructing it using the discarded planks. Some defenders of vague or indeterminate identity claim that: (1) although the reconstructed ship is distinct from the repaired ship, it is indeterminate whether the original ship is the reconstructed ship and indeterminate whether it is the repaired ship, and (2) the indeterminacy is due to the world and not just an imprecision in the language used to describe the situation. I argue that such a description is incoherent. The argument has two features. First, it differs in spirit from Gareth Evans's more general famous proof against the possibility of indeterminate identity. This is because I rely on facts regarding counting and sets. Second, I focus on Terence Parsons's recent defence of indeterminate identity. I argue that his attempts at making sense of counting objects involving indeterminate identities fail on technical and philosophical grounds.