Studies on tree reconstruction often formulate this task as a discrete problem because the tree topology is a discrete object. In this talk, we will discuss an alternative analytic approach to studying the asymptotic theory of tree reconstruction methods. Specifically, we recast the tree reconstruction as a statistical inference problem, where the parameter space is the continuous BHV space of phylogenetic trees. Under this setting, we can use insights from statistical theory to derive analogous results for the tree reconstruction problem.