Grammatical Pattern Recognition For Patients Suffering From Formal Thought Disorder

David Zachary Hafner, Melesio Crespo Sánchez, Edwyn Javier Aldana Bobadilla, Alejandro Molina Villegas

Research output: Contribution to conferencePaperpeer-review


First the de-identified transcripts are edited to remove punctuation markers and to add time stamps to the text. Then we translate words of text into appropriate tags POS (parts of speech) and use them to generate n-grams (with n=2, 3, 4, 5). Then, using computational techniques we cluster the results and obtain the frequencies of different syntax patterns. Our preliminary results suggest that perseveration – whether repetitive stuttering or progressive sentence formation – translate into uncommon syntax patterns. In a pilot study on formal thought disorder, our preliminary results show that between a third and a half of examples of perseveration involve the use of the first person pronoun. This data corresponds with the hypotheses found in the psychoanalytic theories of both Lacan and Bollas that psychotic events involve difficulties in the narration of self. We confirmed Covington et al.’s (2005) results on frequency of different language events, with perseveration and loose associations being more common than phrasal neologisms and direct reference to hallucinations, that were in turn more common than examples of illogicality or grammatical anomalies.
Original languageEnglish
Publication statusPublished - 25 Oct 2019
EventAssociation for the Psychoanalysis of Culture & Society 2019 Annual Conference: Displacement: Precarity & Community - Rutgers University Inn and Conference Center, New Brunswick, United States
Duration: 25 Oct 201927 Oct 2019


ConferenceAssociation for the Psychoanalysis of Culture & Society 2019 Annual Conference
Abbreviated titleAPCS 2019
Country/TerritoryUnited States
CityNew Brunswick
Internet address


Dive into the research topics of 'Grammatical Pattern Recognition For Patients Suffering From Formal Thought Disorder'. Together they form a unique fingerprint.

Cite this