Isn't "backpropagation" just a synonym for the partial derivative of the scalar cost function with respect to the weights? And in the mathematical formulation o

Isn't "backpropagation" just a synonym for the partial derivative of the scalar cost function with respect to the weights? And in the mathematical formulation of biological neural networks these derivatives can't be computed analytically. Your comment sounds like "backpropagation" is some kind of natural phenomenon.