The billions of cells within an individual can be organized by genealogy into a single somatic cell tree that starts from the zygote and ends with present day cells. In theory, this tree can be reconstructed from replication errors that surreptitiously record divisions and ancestry. Such a molecular clock approach is currently impractical because somatic mutations are rare, but more feasible measurements are possible by substituting instead the 5' to 3' order of epigenetic modifications such as CpG methylation. Epigenetic somatic errors are readily detected as age-related changes in methylation, which suggests certain adult stem cells divide frequently and "compete" for survival within niches. Potentially the genealogy of any human cell may be reconstructed without prior experimental manipulation by merely reading histories recorded in their genomes.