Abstract:
Human language has the unique ability to transcend not only the physical world, but also its own literal words. This allows for non-literal “figurative” language such as irony and metaphor, but also carries the risk of misunderstanding. As psychotherapy constitutes the first-line treatment for borderline personality disorder (BPD), the careful choice of words is vital - especially in a disorder characterized by rejection sensitivity, emotional dysregulation, and difficulties in mentalization. However, when it comes to the use of non-literal language, there is a significant disagreement among therapies, and a paucity of empirical data to guide clinical practice. While Dialectical Behavior Therapy (DBT) is rich in metaphor and, in some cases, allows for irony, Mentalization-based therapy strictly discourages their use. And despite the shared symptoms between BPD and schizophrenia, nonliteral language has only been studied in the latter, never in BPD.
Integrating knowledge from both linguistics and psychopathology, the studies in this thesis are the first to test clinical considerations about nonliteral language comprehension in BPD with empirical data. By assessing irony and metaphor, two types of nonliteral language are considered, each controlled for its specific linguistic characteristics. All three studies examine mentalization through cognitive and affective empathy as potential influences on comprehension. Borderline symptoms are assessed dimensionally to address the shortcomings of traditional categorical approaches to personality disorders.
The first study examined metaphor comprehension in patients with BPD compared to healthy controls (HC). The paradigm distinguished between two types of metaphors: novel metaphors, which require active construction of meaning, and conventional metaphors, which are interpreted almost literally. Because conventional expressions may be unfamiliar to the individual, the study also assessed participants' familiarity with them. Patients with BPD had robust comprehension of novel metaphors, suggesting a preserved ability to construct metaphorical meaning. Reported problems with conventional metaphors disappeared when familiarity was taken into account.
Next, the thesis examined irony comprehension in both clinical and non-clinical populations. Social cognition research frequently conflates irony with sarcasm, linking it to a critical attitude and neglecting ironic compliments. With a heightened sensitivity for to rejection, this is particularly problematic for BPD. Thus, the second study devised an irony paradigm that embedded ironic, literal, critical, and praising conversations into a messenger interface. The new paradigm was tested in healthy adults and related to borderline and schizotypal traits. The paradigm showed high reliability and the data confirmed distinct factors for each item category, underscoring the importance of distinguishing ironic praise and criticism. Beyond the known relation between irony and schizotypal symptoms, for the first time, borderline symptoms were associated with reduced irony detection.
The third study aimed to replicate the previous findings on irony detection in BPD patients while controlling for schizotypal symptoms. To achieve this, a signal detection framework was applied to assess sensitivity and response bias. Results confirmed that, independent of schizotypal symptoms, BPD patients were less sensitive than controls in discriminating between ironic and literal remarks. While HC tended to interpret the literalness of a stimulus more critically, BPD patients showed an explicit tendency to perceive praising remarks as less praising.
Across studies, mentalization did not explain performance beyond borderline symptoms, possibly due to its assessment as a self-report questionnaire. Consistent with the continuous conceptualization of personality disorders, borderline symptoms outperformed all group comparisons.
This thesis offers the first empirical evidence on nonliteral language comprehension in BPD. Patients demonstrated comprehension of metaphoric language, but had difficulty with irony , suggesting that it may be the ambiguity of the speaker‘s intention, rather than the nonliteral language itself, that impedes their comprehension. The findings urge caution with irony, but clearly support the use of metaphor in therapy, potentially helping to understand complex contexts, symptoms or feelings, and to establish a common ground for communication.