Files in this item



application/pdfTytko&Augstkalns_2020_ILLS12proceedings.pdf (280kB)


Title:How well do we know ourselves? Identifying suicide markers in online communication: A case study of a graduate student’s writing
Author(s):Tytko, Tetiana; Augstkalns, Mary Catherine
online communication
suicide notes
discourse analysis
suicide markers
Abstract:Linguistic features of suicide notes have been analyzed and compared to forged suicide notes and ordinary letters (Osgood & Walker 1959), diaries, and reports (Baddeley, Daniel, & Pennebaker 2011). However, the real focus should be on determining suicide warning signs in one’s writing prior to the suicide note. Contemporarily, text messages can illustrate informal communication and showcase the emotional state a person experiences at the time of texting (Xie & Kang 2015). This research investigates linguistic features of genuine suicide notes present in the online communication of a suicidal person before attempting suicide. Also, this paper determines common themes discussed in these texts. Personal correspondence of a suicidal person was analyzed over five months. Two corpora (Corpus 1 and Corpus 2) were studied using the following measures: Osgood and Walker’s (1959) allness terms, Hu and Liu’s (2004) negative sentiment lexicon, and English pronouns. AntConc (Anthony 2019) along with the Linguistic Inquiry and Word Count program (Pennebaker, Booth & Francis 2001) were utilized to find frequencies and categorize the person’s psychological states. Findings indicate that the number of allness terms was significantly higher in Corpus 2, immediately prior to the suicide attempt, whereas first-person singular pronouns were equally used in both corpora. Further, Corpus 2 contained more words of positive evaluation which contradicts previous research findings. The discourse analysis of the participant’s messages demonstrated an equal quotient of themes connected to blame discussed in both corpora. These results show that personal writing reflects people’s emotional states better than they think, as the analysis of the texts revealed that themes discussed in Corpus 2 before the suicide attempt were present even five months before, in Corpus 1, which contradicts the participant’s self-report. Therefore, it is the interlocutor’s duty to utilize the tools at hand to assist such people in crisis.
Issue Date:2020
Publisher:Studies in the Linguistic Sciences: Illinois Working Papers
Citation Info:Studies in the Linguistic Sciences: Illinois Working Papers 43: 45-62.
Date Available in IDEALS:2020-12-01

This item appears in the following Collection(s)

Item Statistics