The bei-construction is one of the most important yet difficult grammar points taught in Chinese Second Language (L2) curricula. Previous studies have discussed its structural complexity and the learning difficulties revealed by interlanguage data. However, syntactical descriptions alone are not sufficient to address the challenge that learners face in grasping why and when to use this construction. My research addresses this challenge by using Halliday's framework of language functions and proposing a generalized description of the construction's function. As I show, the widely accepted "Adversity Hypothesis" is inadequate to describe the kinds of contexts that require the bei-construction. Based on my analysis of contemporary spoken Mandarin from the data I collected, this dissertation finds that the essential function of the bei-construction is to express nonvolitionality, and this generalization provides a unifying explanation that accounts for its logical, textual, and interpersonal functions. In order to address the kinds of problems students of Mandarin face when learning the bei-construction, this dissertation also identifies the errors that Chinese L2 learners make when using it. These errors occur in the interlanguage data, and analysis of them reconfirms that the construction's function requires more pedagogical attention. However, my evaluation of Chinese L2 textbook samples demonstrates that the current pedagogical design fails to reflect how native speakers use the bei-construction and fails to address the difficulties in learning its functions. In order to fill this gap, this dissertation proposes a sequence in which the function of the bei-construction should be taught in Chinese L2 curricula within the framework of pedagogical grammar.