In the last decade, we have seen emerging research exploring the use of natural language processing (NLP) techniques for assisting in the identification of clinical conditions that affect language. One of these clinical conditions is language impairment, a disorder identified by delayed or disordered language patterns in an individual with normal intelligence with no neurological or other physiological conditions. In this article, we present a survey of this emerging line of research, which for the most part has focused on the task of discriminating the clinical from the non-clinical group by posing the task as an automated classification problem. The focus of this survey is on the types of features recent research has explored. We also discuss the many interesting open questions that this research has yet to explore.