By reviewing the history of our understanding of several related graphs with the elements of E or Xie in the recently excavated Warring States manuscripts, this article discusses two different approaches to paleographic decipherment of the Warring States Chu script. The author argues that in the decipherment of the Warring States script, in which there recorded varieties of sounds and speeches and multiplicities of graphic variants, the method of decipherment should also shift from the previously graph-centered analysis to a combination of the phonetic, graphic, and semantic analysis. The author points out the pitfalls in the current scholarly literature on paleographic decipherment, and suggests a more rigorous phonological approach to paleographic decipherment.