Psychological disorders, especially internalizing disorders like anxiety and depression, cause immense human and economic burden across the globe. Prior work shows that internalizing disorders are characterized by perturbations in emotion regulation (i.e., the strategies people use to change how they feel), with excessive use of maladaptive strategies that reduce short-term distress but maintain long-term impairment and insufficient use of adaptive strategies that allow individuals to escape these cycles of impairment. Developing tools that identify poor emotion regulation and improve this critical affective skill could help address the global burden of psychopathology.
My dissertation approaches this problem by focusing on a potential linguistic signature of effective emotion regulation, called linguistic distancing. People can use several strategies to regulate their emotions, including “psychological distancing,” in which one thinks about negative situations as separated or removed from oneself, for instance by replaying memories from a third-person perspective. Linguistic distancing involves using language to achieve this psychological shift, for instance by eliminating use of first-person singular pronouns or present-tense verbs. My dissertation tests (i) whether this linguistic strategy is indeed associated with effective emotion regulation, (ii) whether use of linguistic distance varies across age, and (iii) the potential translational value of linguistic distancing in assessing symptom severity and treatment outcomes in psychotherapy.
The general introduction of the dissertation integrates theoretical and empirical research supporting relations between linguistic distancing, emotion regulation, and mental health. I review both foundational and contemporary studies on these topics, arriving at a set of clear hypotheses for the dissertation as a whole. For example, I draw upon neuroscientific perspectives on the neural representation of psychological distance as evidence for the notion that shifting pronoun and verb use should facilitate more distant psychological representations and consequently down-regulate negative affect.
Paper 1 (Nook, Schleider, & Somerville, 2017, Journal of Experimental Psychology: General) then reports on two studies demonstrating that linguistic distancing shares a bidirectional relationship with emotion regulation: Linguistic distancing both tracks successful cognitive reappraisal and can be used to down-regulate negative affect. The first of these studies (N = 107) and its replication study (N = 121) asked participants to transcribe their thoughts and feelings while either responding naturally to aversive images or while using cognitive reappraisal to regulate their reactions to these images. Linguistic analyses revealed that participants spontaneously used fewer first-person singular pronouns and present-tense verbs (established markers of linguistic distancing) when regulating their emotions. Further, we found that how strongly one distanced one’s language when reappraising correlated with regulatory success. The second study (N = 227) and its replication (N = 247) reversed this relationship and showed that merely distancing one’s language (i.e., asking participants to write about negative images without using the word “I” or present-tense verbs) spontaneously reduced their negative affect. Indeed, participants reported feeling less negative affect when they made these subtle linguistic shifts. Taken together, this paper provides replicable evidence that linguistic distancing can both measure and manipulate emotion regulation: Greater spontaneous linguistic distancing tracks stronger reappraisal success, and merely distancing one’s language can down-regulate emotion.
Paper 2 (Nook*, Vidal-Bustamante*, Cho, & Somerville, 2020, Emotion) takes a developmental perspective, asking whether the strength of linguistic distancing during emotion regulation might vary across childhood, adolescence, or young adulthood. In this preregistered study (N = 112), participants aged 10-23 completed the cognitive reappraisal task of Paper 1. Although we once again found that linguistic distancing increased during reappraisal and correlated with successful regulation, we did not find that either linguistic distancing or reappraisal success varied across age. As such, even as early as age 10, spontaneous linguistic distancing tracks successful emotion regulation.
Paper 3 (Nook, Hull, Nock, & Somerville, 2022, Proceedings of the National Academy of Sciences) takes a translational perspective, providing evidence that linguistic distance is related to the severity of clients’ internalizing symptoms and their treatment outcomes in a large real-world corpus of therapy transcripts. We computed measures of linguistic distancing within a corpus of > 1.2 million text samples from 6,229 clients as they used a messaging-based psychotherapy service. We split this sample into exploratory (N = 3,729) and validation (N = 2,500) subsets, and preregistered all analyses before replicating them in the validation dataset. We found that internalizing symptoms (i.e., depression and anxiety) decreased over time in therapy and that linguistic distancing correlated with symptom severity at both within-person and between-person levels. This means that patients reported fewer internalizing symptoms on months in which they used higher linguistic distance. As such, not only does linguistic distance serve as a measure of emotion regulation in the lab, it also assesses actual mental health symptoms in real-world therapy transcripts.
Finally, the general discussion synthesizes results from this dissertation to articulate the basic and translational impacts of research on linguistic distancing. This leads to the formulation of an overarching theoretical model in which taking a distanced perspective on one’s habitual thoughts, feelings, and behaviors might be a key psychological process in fostering mental health. Language may be one path for both measuring how strongly individuals habitually take this distanced perspective and a tool for facilitating this shift. Potential future directions for testing this model and addressing open questions raised by this dissertation are discussed.
Evidence from this dissertation supports the notion that linguistic distance is a marker of effective emotion regulation and mental health across childhood, adolescence, and young adulthood. The studies consistently use Open Science practices of preregistration, replication, and data sharing to add certainty and transparency to results. My work is also intentionally interdisciplinary, integrating tools from affective science, developmental science, psycholinguistics, and clinical psychology. This approach yields findings that I believe address open questions in the field, extend existing theory, and have the potential to stimulate future applied research that uses linguistic distancing as a tool for detecting and intervening on psychological symptoms at a large scale (e.g., online or in ecological momentary text sampling). In all, I hope that this line of research meaningfully contributes to scientific understanding of emotion and germinates additional work that can foster emotional health.
