Pratik Bhandari

Studies of human language processing employ metrics like cloze probability and surprisal as proxies of a word's contextual probability. There already exist corpora of cloze probability-normed sentences for many languages of Germanic and Romance families, but relatively few resources are available in other language families, and none whatsoever in the Nepali language. Here we report a dataset of cloze-normed Nepali sentences together with the information-theoretic measure of surprisal. We also calculate the contextual constraint of the sentences using metrics of type-token ratio and proportion of idiosyncratic responses. We establish the relationships between a word's contextual probability and the contextual constraint of the sentence generating those words. With this, the present study aims to create a corpus of cloze-normed Nepali sentences, providing resources for diverse research (e.g., predictive language processing) in this understudied language.

Introduction: Hearing loss is a global health problem. It is not only an impairment of hearing but also has debilitating effects on the social, emotional and cognitive domains. It also imposes costs on a country’s healthcare sector, which must first be identified to prevent the prevalence of the impairment. In a low- and middle-income country like Nepal, there is no published systematic review on the prevalence of hearing loss. We aim to perform a systematic review and meta-analysis on the national prevalence of hearing loss in Nepal across all age groups. Methods and analysis: We will report our findings in line with the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols guidelines. We will search the MEDLINE (PubMed) and Nepal Journals Online databases for publications in Nepali and English. Google Scholar will also be used to broaden the search scope. We will also search grey literature, like government reports. Studies that report the prevalence of hearing loss in Nepal will be included. Titles and abstracts of the studies will be independently screened by both reviewers (PB and SA) using Rayyan software. Full-text articles that pass the initial screening will also be independently reviewed by both reviewers. Disagreements will be resolved through discussion at all stages. If necessary, a third reviewer will be invited, and the original author will be contacted where appropriate (e.g., when data are missing in the original study). We will perform a narrative synthesis of the study results. Cochran’s Q test and the I² statistic will be calculated to assess heterogeneity. I² values of 25%, 50% and 75% will be interpreted as indicating low, moderate and high heterogeneity, respectively. We will conduct a meta-analysis using a random-effects model, which accounts for both within-study and between-study variability. We will perform a sensitivity analysis to examine whether a single study disproportionately contributes to heterogeneity. Potential publication bias will be evaluated using both visual and statistical approaches (Funnel plot and Egger’s test). In addition, if funnel plot asymmetry is observed, trim-and-fill analysis will be conducted. Statistical tests will be performed in R (V.4.3.1) using RStudio (V.2023.06.1+524) with the metafor package for meta-analysis. Ethics and dissemination: Ethics approval is not required for this protocol. This review and meta-analysis are based entirely on secondary data; no primary data will be collected. Our findings will be useful to quantify the national prevalence of hearing impairment in Nepal. They may be useful for developing intervention policies and furthering research agendas for government agencies and hearing-health researchers alike. The results will be presented at conferences, published in peer-reviewed journals and shared with relevant policy-making bodies. PROSPERO registration number: CRD420251120903.

Previous research has shown that semantic predictability facilitates language comprehension at a moderate level of spectral degradation (e.g., 4-channel noise-vocoding). Clean speech and self-paced reading studies have shown that contextual facilitation decreases with an increase in presentation rate (fast speech), while the previous evidence on the enhancement of contextual facilitation with a decrease in presentation rate (slow speech) is inconsistent. We aimed to examine how a change in speech rate modulates contextual facilitation in language comprehension when the speech is moderately degraded at 4-channel noise-vocoding. In two experiments, we compared participants’ word recognition in a sentence while they listened to a moderately degraded speech presented at normal and fast speech rates (Experiment 1), and at normal and slow speech rates (Experiment 2). We also varied the predictability of the sentence-final target-word (high and low predictability). Results of the study demonstrated that fast speech limits the time for lexical processing. This time constraint interferes with the lexical processing of words, disproportionately affecting the low predictability sentences, on top of effortful listening of the moderately degraded speech. In contrast, slow speech does not amplify the contextual facilitation; lexical processing, context representation, and semantic predictions are optimal at the normal speech rate for a moderately degraded speech.

The aim of this study was to examine the role of attention in understanding linguistic information even in a noisy environment. To assess the role of attention, we varied task instructions in two experiments in which participants were instructed to listen to short sentences and thereafter to type in the last word they heard or to type in the whole sentence. We were interested in how these task instructions influence the interplay between top-down prediction and bottom-up perceptual processes during language comprehension. Therefore, we created sentences that varied in the degree of predictability (low, medium, and high) as well as in the degree of speech degradation (four, six, and eight noise-vocoding channels). Results indicated better word recognition for highly predictable sentences for moderate, though not for high, levels of speech degradation, but only when attention was directed to the whole sentence. This underlines the important role of attention in language comprehension.

Previous studies have shown that at moderate levels of spectral degradation, semantic predictability facilitates language comprehension. It is argued that when speech is degraded, listeners havenarrowed expectationsabout the sentence endings; i.e., semantic prediction may be limited to only most highly predictable sentence completions. The main objectives of this study were to (i) examine whether listeners form narrowed expectations or whether they form predictions across a wide range of probable sentence endings, (ii) assess whether the facilitatory effect of semantic predictability is modulated by perceptual adaptation to degraded speech, and (iii) use and establish a sensitive metric for the measurement of language comprehension. For this, we created 360 German Subject-Verb-Object sentences that varied in semantic predictability of a sentence-final target word in a graded manner (high, medium, and low) and levels of spectral degradation (1, 4, 6, and 8 channels noise-vocoding). These sentences were presented auditorily to two groups: One group (n=48) performed a listening task in an unpredictable channel context in which the degraded speech levels were randomized, while the other group (n=50) performed the task in a predictable channel context in which the degraded speech levels were blocked. The results showed that at 4 channels noise-vocoding, response accuracy was higher in high-predictability sentences than in the medium-predictability sentences, which in turn was higher than in the low-predictability sentences. This suggests that, in contrast to thenarrowed expectationsview, comprehension of moderately degraded speech, ranging from low- to high- including medium-predictability sentences, is facilitated in a graded manner; listeners probabilistically preactivate upcoming words from a wide range of semantic space, not limiting only to highly probable sentence endings. Additionally, in both channel contexts, we did not observe learning effects; i.e., response accuracy did not increase over the course of experiment, and response accuracy was higher in the predictable than in the unpredictable channel context. We speculate from these observations that when there is no trial-by-trial variation of the levels of speech degradation, listeners adapt to speech quality at a long timescale; however, when there is a trial-by-trial variation of the high-level semantic feature (e.g., sentence predictability), listeners do not adapt to low-level perceptual property (e.g., speech quality) at a short timescale.