Explain why the sort of learning rule discussed in this chapter is appropriately called “backpropagation.”