Previous linguistic research suggests that warning behavior of violent intent, as established in the field of psychology (Reid Meloy et al, 2011, p. 265), can be found textually using computational text analysis tools (Brynielsson et al, 2013, p. 3; Cohen et al, 2013, p. 248). Only “Leakage” warning behavior, which is the release of indications of ill intent to a third party, has been linguistically explored (Kaati et al, 2016, p. 2). This study will compile the personal documents of school shooters into a corpus and search the corpus for comparable frequencies of categories of language as defined by the text analysis program LIWC (Linguistic Inquiry and Word Count). However, manual data collection will be used instead. The use of 3rd person plural pronouns will be investigated in this study. This study will search for previously proposed but never explored linguistic indicators of “Leakage”, which is a high frequency of auxiliary verbs paired with verbs that indicate violent action or intent. The resulting frequencies will be compared between those found within the school shooter documents and those found on the base-rate documents. These documents are online blogs and the texts of other known offenders known as “lone wolves”, defined as individuals that act of their own accord with no relation to a group (Brynielsson et al, 2013, p. 3). The goal of this study is to find similar frequencies in the texts of school shooters as that of other “lone wolf” offenders to corroborate past linguistic research, as well as to determine whether these frequencies are relevant enough to be used in finding future perpetrators.