How can routine cell replacement permit psychological continuity?

If every seven years the human body has a renewed set of cells, and if memory is based on structural artifacts, then how come we remember old memories? Parallel distributed processing, coupled with adult neurogenesis and synaptogenesis, provide the necessary redundancy for handling this continuous neuron replacement. By cross-referencing sparse representations, psychological continuity easily bridges episodes of isolated cell death. This long-term progressive aspect is generally ignored in machine learning.