Linguistic agentivity as a marker of traumatic discourse

Abstract

Objectives

During reification in the face of an aggressor, helplessness in the face of a natural disaster or the death of a loved one, or inability to flee political violence, the notion of loss of agency proves to be consubstantial with traumatic confrontations. The cardinal post-traumatic symptoms in themselves increase this loss of the sense of agency: uncontrollable flashbacks, extensive avoidances, depression that limits capacities for action, depersonalization. The psychotraumatic consequences sometimes take control of a large part of one's existence, even if it means inducing new traumas. The aim of this study was thus to examine how psychologically injured people encode agentivity in their traumatic discourses. Agency refers to the fact that humans can make choices that affect their existence or their environment, choices that produce changes. Several studies have shown that people with post-traumatic stress disorder (PTSD) have a diminished sense of agency. We therefore hypothesized that, linguistically speaking, subjects suffering from PTSD would use more stative verbs and fewer action verbs than controls.

Materials and methods

Three corpora of trauma narratives produced by victims of the Paris attacks that took place in 2015, as well as the narratives of French soldiers returning from the war in Afghanistan, were compared with the narratives of a control group. The proportions of different types of static verbs and action verbs used with the first-person pronoun “I” were analysed. We thus detailed the uses of verbs of quality, existence, locations, cognitions, and perceptions, with a subdivision of action verbs into verbs of action, displacement, and movement.

Results

Our results show that traumatic discourses contained significantly fewer action verbs, particularly verbs denoting movement, in favour of static verbs, particularly verbs of cognition and perception, especially in the Paris attacks corpus. As for action verbs, the control group and to a lesser extent the Afghanistan group used more action and displacement verbs, while the immediate Bataclan group produced more movement verbs.

Discussion

As we suspected, traumatic speech was less prone to action verbs, this linguomarker reflecting a low sense of agency. The identifiable differences between the corpora of psychologically injured subjects can be linked to the traumatic circumstances (i.e., the place was closed after the Paris attacks) and to the populations studied (e.g., military staff engaged in combat action). These data confront us with the need to define not only clinical and psychometric sub-phenotyping of post-traumatic disorders, but also psycho-linguotyping according to certain variables: age of occurrence during the event, individual or collective trauma, temporality of exposures (single trauma, short, long, or repeated), etcetera.

Conclusions

The proportion of stative versus active verbs contributed to the diagnosis of PTSD, including these phenotypic expressions. This linguomarker, among other purposes, might be used to guide effective speech acts in practitioner–patient interlocution and, more generally, to evaluate the effectiveness of recommended treatments. The use of the linguomarker must therefore be understood as the mobilization of a neurophysiological function, making it possible to move from traumatized speech to protective speech.