Abstract
Objective: Although language use in schizophrenia has been widely studied, the pauses occurring during patients’ speech have been the subject of relatively few investigations. This study examined the frequency of discourse markers functioning as pause fillers in the elicited spontaneous speech of individuals with schizophrenia.

Methods: The study included 38 patients diagnosed with schizophrenia (SCH) and 40 healthy controls. To obtain elicited spontaneous speech samples, participants were asked to speak freely for 10 minutes on a given topic. After transcribing the recordings, the frequency of five selected discourse markers (hani, yani, şey, işte, falan) functioning as pause fillers was analyzed, and a total Filler Word (FW) index was calculated. The Positive and Negative Syndrome Scale (PANSS) was administered to the schizophrenia group to assess symptom severity.

Results: The total FW index was significantly lower in the SCH group than in controls in the unadjusted comparison (M = 4.48, SD = 2.99 vs. M = 5.76, SD = 2.61; t = −2.010, p = .048), but this difference was no longer significant after sex was included in the model. In analyses including sex as a factor, controls showed significantly higher “Hani” and “Falan” indices than patients with schizophrenia (p < .05). No FW index was significantly correlated with PANSS subscale or total scores (all p > .05). In binary logistic regression controlling for age and sex, only the “Hani” index significantly predicted group membership, with higher scores associated with increased odds of belonging to the control group (OR = 4.987, 95% CI [1.399, 17.773], p = .013).

Conclusion: Individuals with schizophrenia differed from healthy controls in their use of discourse markers functioning as pause fillers, suggesting that spontaneous speech may contain subtle linguistic features observed in schizophrenia. The lack of association with symptom severity suggests that these features may reflect broader pragmatic and discourserelated language characteristics rather than current illness severity. Future longitudinal studies with larger samples are needed to further evaluate the clinical relevance of these preliminary findings.

Key words: Discourse markers, Filler words, Illness severity, Schizophrenia, Speech