Linguistic Research Research Articles

How ready is speech-to-text for psychological language research? Evaluating the validity of AI-generated English transcripts for analyzing free-spoken responses in younger and older adults.

For the longest time, the gold standard in preparing spoken language corpora for text analysis in psychology was using human transcription. However, such standard comes at extensive cost, and creates barriers to quantitative spoken language analysis that recent advances in speech-to-text technology could address. The current study quantifies the accuracy of AI-generated transcripts compared to human-corrected transcripts across younger (n = 100) and older (n = 92) adults and two spoken language tasks. Further, it evaluates the validity of Linguistic Inquiry and Word Count (LIWC)-features extracted from these two kinds of transcripts, as well as transcripts specifically prepared for LIWC analyses via tagging. We find that overall, AI-generated transcripts are highly accurate with a word error rate of 2.50% to 3.36%, albeit being slightly less accurate for younger compared to older adults. LIWC features extracted from either transcripts are highly correlated, while the tagging procedure significantly alters filler word categories. Based on these results, automatic speech-to-text appears to be ready for psychological language research when using spoken language tasks in relatively quiet environments, unless filler words are of interest to researchers.

Adaptive Self-Reflection as a Social Media Self-Effect: Insights from Computational Text Analyses of Self-Disclosures of Unreported Sexual Victimization in a Hashtag Campaign

Hashtag campaigns calling out sexual violence and rape myths offer a unique context for disclosing sexual victimization on social media. This study investigates the applicability of adaptive self-reflection as a potential self-effect from such public disclosures of unreported sexual victimization experiences by analyzing 92,583 tweets that invoked #WhyIDidntReport. A supervised machine learning classifier determined that 61.8% of the tweets were self-disclosures of sexual victimization. Linguistic Inquiry and Word Count (LIWC) analysis showed statistically significant differences in four psycholinguistic dimensions (greater use of past focus, cognitive processes, insight, and causation words) connected with reflective processing in tweets with self-disclosed sexual victimization compared to those without. Additionally, topic modeling and thematic analysis identified nine salient topics within the self-disclosing tweets, comprising three self-distanced representations (i.e., relatively abstract and insightful construals) of the unwanted experiences: (a) acknowledging one’s previously unacknowledged victimization, (b) reaffirming one’s rationale for not reporting, and (c) decrying invalidating response to one’s disclosure. Moving beyond reception effects and social support in extant research about social media as a coping tool, this study provides new empirical insights into the potential of social media to promote expressive meaning-making of upsetting and traumatic experiences in ways that support recovery and resilience.

Register and the dual nature of functional correspondence: accounting for text-linguistic variation between registers, within registers, and without registers

Abstract During the past 20 years, corpus linguistic research on register variation has yielded important theoretical advances. The first part of this paper discusses these advances and the cumulative body of research that has produced them. In the second part of the paper, we focus on the goals of research on register variation. The traditional goal of the text-linguistic (TxtLx) approach to linguistic variation has been to describe registers and patterns of register variation: describing the linguistic and situational characteristics of registers. In this paper, we explore a related, but distinct, text-linguistic goal: to account for all linguistic variation among texts. Because the TxtLx framework assumes the importance of functional correspondence between linguistic characteristics and situational characteristics, it is reasonable to assume that in addition to register, we can use situational parameters coded continuously at the level of individual texts as additional predictors of text-linguistic variation. We describe the results of an empirical study to show that using both register categories and text-level situational parameters as predictors results in a more comprehensive and explanatory model of text-linguistic variation. In the conclusion we discuss the future of corpus-based register studies, focusing on unanswered questions related to theoretical claims about register.

Cultural Concepts in the Turkish Translation of „The Knight in the Panther’s Skin“

The present work is a critical analysis of Shota Rustaveli's poem "The Knight in the Panther’s Skin" translated into Turkish by Professor Bilal Dindar and Assistant Professor Zeinelabidin Makas. We will present comparison of the original and translated texts of "The Knight in the Panther’s Skin". "The Knight in the Panther’s Skin" is one of the remarkable monuments of medieval literature. "The Knight in the Panther’s Skin" is an epic work containing lyrical passages, the story of which takes place in Arabia and India. We often find cultural-specific concepts in the target language, the translator tries to adapt the lexical units in the original text to his own culture. These cultural concepts are: sun, moon, devil, wine drinking. We used the following methods during the research: Corpus Linguistic research method, which is relevant in modern humanitarian studies. The use of this method allows to quickly and efficiently process a large amount of empirical material, which simplifies the research process. We have also used method of linguoculturological analysis. Modern linguoculturology studies are of a comparative-contrastive nature, because the national specificity and universality of the concepts are particularly clearly revealed when comparing the materials of two or more languages. According to the linguocultural analysis, we will study the cultural concepts in two different cultural texts, which will help in the detailed description of the cultural concepts expressed in the linguistic process.

Analysing the Impact of Artificial Intelligence on the Development of Contemporary Philology: The Use of Automated Tools in Linguistic Research

The relationship between linguistics and artificial intelligence is still controversial. The relevance of this paper lies in the interpretation of technological progress as promising for the field of linguistics, since the synergy between artificial intelligence and linguistics can lead to a rethinking of linguistic paradigms. This paper is methodologically based on a detailed study of the related literature on the recent history of these two fields. As a result, the paper offers a detailed analysis of the related scientific literature and revises the relationship between linguistics and artificial intelligence. Artificial intelligence is seen as a branch of human science, and models have emerged in the intermediate period between the two fields, which were not always understood, but acted as buffers against the interaction of one field with the other; if artificial intelligence is a branch of human science, then linguistics can also adopt this heritage, which may lead to a rethinking of the paradigms of linguistics and challenge the opposition between the exact and experimental sciences. The evolution of the ways of thinking about automatic language processing from the cognitive science point of view of artificial intelligence is traced, and different types of knowledge considered in linguistic computing are discussed. The importance of language both in human-machine communication and in the development of reasoning and intelligence is emphasised. Ideas that can overcome the internal contradiction of artificial intelligence are put forward. The practical significance of the paper is that it highlights how, with the help of automatic language processing, humanity has gradually realised the need to use technological knowledge to understand language. This combination opens up not only knowledge about the language itself, but also general knowledge about the world, about the speaker’s culture, about the communication situation, and about the practice of human relations. The conclusions of the study point to the great potential of interaction between linguistics and artificial intelligence for the further development of both fields.

Emotional tone in clinical high risk for psychosis: novel insights from a natural language analysis approach.

Individuals at clinical high risk (CHR) for psychosis experience subtle emotional disturbances that are traditionally difficult to assess, but natural language processing (NLP) methods may provide novel insight into these symptoms. We predicted that CHR individuals would express more negative emotionality and less emotional language when compared to controls. We also examined associations with symptomatology. Participants included 49 CHR individuals and 42 healthy controls who completed a semi-structured narrative interview. Interview transcripts were analyzed using Linguistic Inquiry and Word Count (LIWC) to assess the emotional tone of the language (tone -the ratio of negative to positive language) and count positive/negative words used. Participants also completed clinical symptom assessments to determine CHR status and characterize symptoms (i.e., positive and negative symptom domains). The CHR group had more negative emotional tone compared to healthy controls (t=2.676, p=.009), which related to more severe positive symptoms (r2=.323, p=.013). The percentages of positive and negative words did not differ between groups (p's>.05). Language analyses provided accessible, ecologically valid insight into affective dysfunction and psychosis risk symptoms. Natural language processing analyses unmasked differences in language for CHR that captured language tendencies that were more nuanced than the words that are chosen.

MACHINE LEARNING AND THE STUDY OF LANGUAGE CHANGE: A REVIEW OF METHODOLOGIES AND APPLICATION

Language change, a fundamental aspect of human communication, has long been a central focus in linguistic research. Traditional methods of analysis, while valuable, have been limited by the scale and complexity of linguistic data. The advent of machine learning (ML) offers transformative potential in this field, enabling the analysis of vast datasets and the discovery of subtle patterns that may elude manual scrutiny. This review paper comprehensively examines the current state of ML methodologies in the study of language change, synthesizing findings from 67 peer-reviewed articles. We delve into diverse ML approaches, including supervised, unsupervised, and deep learning techniques, and critically evaluate their applications across various linguistic domains, such as historical linguistics, sociolinguistics, and language contact. We address challenges related to data availability, bias, and model interpretability, emphasizing the need for transparent and rigorous methodologies. By summarizing key findings and outlining future directions, this review aims to foster interdisciplinary collaboration between linguists and computer scientists, advancing our understanding of the complex dynamics of language evolution.

A Linguistic Analysis of United States Navy Orthopaedic Surgery Applicant Personal Statements.

Despite the importance of linguistic analysis, no systematic research has been explored in the form of linguistic analysis on personal statements for military orthopedic surgery residency programs. This study was conducted to analyze U.S. Navy (USN) orthopedic surgery applicants' personal statements using an automated textual analysis program to assess personal statements for linguistic styles. A retrospective analysis of USN orthopedic applicant personal statements from application years 2016 to 2019 was performed utilizing the Linguistic Inquiry and Word Count (LIWC) software. LIWC analyzed the text for summary variables: analytical thinking, clout, authenticity, and emotional tone. We compared this analysis with Step 1 and Step 2 scores and determined whether an applicant matched. A total of 94 personal statements (60,230 words) were analyzed using LIWC. The average word count was 640.7, with an average of 23 words per sentence. The average-matched applicant USMLE Step 1 and Step 2 scores were 240 and 250, respectively. When examining summary traits utilizing multiple logistic regression analysis, only analytical thinking demonstrated a statistically significant difference in matched versus unmatched applicants with a P = .011 (OR = 1.10). As the USMLE Step 1 exam transitions from a scoring system to Pass/Fail grading, programs will look at other characteristics to determine who would likely succeed in residency. From a linguistic analysis standpoint, matched applicants' personal statements demonstrated higher analytical thinking, clout, affiliation, power, and risk focus than unmatched applicants. Unmatched applicants demonstrated higher authenticity than matched applicants.

Hierarchical syntax model of music predicts theta power during music listening

Linguistic research showed that the depth of syntactic embedding is reflected in brain theta power. Here, we test whether this also extends to non-linguistic stimuli, specifically music. We used a hierarchical model of musical syntax to continuously quantify two types of expert-annotated harmonic dependencies throughout a piece of Western classical music: prolongation and preparation. Prolongations can roughly be understood as a musical analogue to linguistic coordination between constituents that share the same function (e.g., ‘pizza’ and ‘pasta’ in ‘I ate pizza and pasta’). Preparation refers to the dependency between two harmonies whereby the first implies a resolution towards the second (e.g., dominant towards tonic; similar to how the adjective implies the presence of a noun in ‘I like spicy … ’). Source reconstructed MEG data of sixty-five participants listening to the musical piece was then analysed. We used Bayesian Mixed Effects models to predict theta envelope in the brain, using the number of open prolongation and preparation dependencies as predictors whilst controlling for audio envelope. We observed that prolongation and preparation both carry independent and distinguishable predictive value for theta band fluctuation in key linguistic areas such as the Angular, Superior Temporal, and Heschl's Gyri, or their right-lateralised homologues, with preparation showing additional predictive value for areas associated with the reward system and prediction. Musical expertise further mediated these effects in language-related brain areas. Results show that predictions of precisely formalised music-theoretical models are reflected in the brain activity of listeners which furthers our understanding of the perception and cognition of musical structure.

Differing Content and Language Based on Poster-Patient Relationships on the Chinese Social Media Platform Weibo: Text Classification, Sentiment Analysis, and Topic Modeling of Posts on Breast Cancer.

Breast cancer affects the lives of not only those diagnosed but also the people around them. Many of those affected share their experiences on social media. However, these narratives may differ according to who the poster is and what their relationship with the patient is; a patient posting about their experiences may post different content from someone whose friends or family has breast cancer. Weibo is 1 of the most popular social media platforms in China, and breast cancer-related posts are frequently found there. With the goal of understanding the different experiences of those affected by breast cancer in China, we aimed to explore how content and language used in relevant posts differ according to who the poster is and what their relationship with the patient is and whether there are differences in emotional expression and topic content if the patient is the poster themselves or a friend, family member, relative, or acquaintance. We used Weibo as a resource to examine how posts differ according to the different poster-patient relationships. We collected a total of 10,322 relevant Weibo posts. Using a 2-step analysis method, we fine-tuned 2 Chinese Robustly Optimized Bidirectional Encoder Representations from Transformers (BERT) Pretraining Approach models on this data set with annotated poster-patient relationships. These models were lined in sequence, first a binary classifier (no_patient or patient) and then a multiclass classifier (post_user, family_members, friends_relatives, acquaintances, heard_relation), to classify poster-patient relationships. Next, we used the Linguistic Inquiry and Word Count lexicon to conduct sentiment analysis from 5 emotion categories (positive and negative emotions, anger, sadness, and anxiety), followed by topic modeling (BERTopic). Our binary model (F1-score=0.92) and multiclass model (F1-score=0.83) were largely able to classify poster-patient relationships accurately. Subsequent sentiment analysis showed significant differences in emotion categories across all poster-patient relationships. Notably, negative emotions and anger were higher for the "no_patient" class, but sadness and anxiety were higher for the "family_members" class. Focusing on the top 30 topics, we also noted that topics on fears and anger toward cancer were higher in the "no_patient" class, but topics on cancer treatment were higher in the "family_members" class. Chinese users post different types of content, depending on the poster- poster-patient relationships. If the patient is family, posts are sadder and more anxious but also contain more content on treatments. However, if no patient is detected, posts show higher levels of anger. We think that these may stem from rants from posters, which may help with emotion regulation and gathering social support.

Researchers Training Researchers: Ethics Training in Quantitative Applied Linguistics

AbstractThis forum piece reports on a brief survey of current quantitative research ethics training materials in Applied Linguistics (AL). This survey was conducted as a step towards an increased understanding of the ways in which we, as a field, train students and researchers to conduct quantitative research ethically. The survey was carried out in the process of creating research ethics training materials as part of a grant‐funded project related to questionable research practices (QRPs). Through manual and computer‐assisted searches in twenty‐four textbooks and twenty‐three course syllabi from the past five years, three themes emerged: (1) research ethics is most often conceptualized in terms of IRB‐related topics, (2) when ‘ethical gray‐zone’ issues are mentioned, the focus is primarily on methodological transparency, data sharing/open science, and selecting the appropriate research design and statistical tests, and (3) materials tend to address the topic of research ethics in a single section or chapter, or in a single day or week of a course. Against this background, we provide three recommendations that our field can implement to provide more robust and thorough research ethics training for students and researchers in TESOL and Applied Linguistics: (1) expand our conceptualization of research ethics to include ethical gray‐zone issues, (2) be thorough and explicit in our discussion of ethical issues related to research decisions, and (3) incorporate research ethics as a recurring theme throughout textbooks and courses.

Linguistic-induced life insurance consumption

Linguistic research shows that there are differences in how language distinguishes future events from the present. When a language speaker separates the grammatical structure of a language from the future and present, they also separate the future and present in behavior when facing intertemporal choices. In this study, we investigate language induced time perception as a determinant of cross-country life insurance consumption. The evidence demonstrates that life insurance consumption is higher in countries where there is non-obligatory grammatical future tense in language. This result still confirms the language-insuring hypothesis after considering cultural dimensions, alternative language variables, alternative insurance variable and endogeneity problem. To sum up, the results suggest that language featured by future time reference affects speaker’s time perception and life insurance decision. Our evidence also provides implication for life insurer in multinational businesses.

Linguistic Analysis of Nonsuicidal Self-Injury Reddit Posts: Implications for Family Therapy

Nonsuicidal self-injury (NSSI) is a growing phenomenon, yet there is still a limited understanding of the behavior, intent behind the behavior and what individuals themselves say about their behavior. This study collected pro-NSSI public blog posts from Reddit and analyzed the content linguistically using Linguistic Inquiry and Word Count software, to examine the use of NSSI specific words, linguistic and psychological linguistic properties. The results inform current mental health practices by dispelling myths and providing insight into the inner world of people who engage in NSSI. The linguistic properties found in the analysis reflected the predicted results; authors of pro-NSSI posts used first-person singular pronouns extensively, indicating high levels of mental health distress and isolation. This study demonstrates how understanding the language of self-harm can help inform mental health treatment using family therapy.

Exploring Stemming Techniques in Ambon Malay Languages: A Systematic Literature Review

Stemming in Ambonese posed a significant challenge due to its extensive lexicon, encompassing approximately 127,000 base words as recorded in the Kamus Besar Bahasa Indonesia (Indonesian Dictionary). This complexity arises from the task of extracting base words from those with affixes, necessitating the removal of various affixes such as prefixes, infixes, suffixes, and their combinations. This process greatly influences analytical outcomes. To address this linguistic complexity, several stemming algorithms were developed. These include Nazief Adriani, Enhanced Confix Stripping, Sastrawi, and Tala, each offering unique techniques to handle stemming complexities in Indonesian. The selection of the appropriate algorithm is crucial for ensuring the accuracy and reliability of the stemming process within the analytical framework. In conducted stemming research, there were variations in methods used. The most frequently used algorithm was Nazief Adriani, with 17 recorded cases, followed by Enhanced Confix Stripping with 12 cases. Sastrawi, although less frequent, was used in 4 cases, while Tala appeared in 1 case. This diversity reflects the available choices in selecting a fitting stemming method. However, this may relate to factors such as ongoing research projects, funding availability, or other external conditions affecting research production during that period. Consequently, stemming research remains an interesting and relevant topic, with the potential for continued growth and significant contributions to text processing and linguistic research in the future.

Integrated End-to-End automatic speech recognition for languages for agglutinative languages

The relevance of the problem of automatic speech recognition lies in the lack of research for low-resource languages, stemming from limited training data and the necessity for new technologies to enhance efficiency and performance. The purpose of this work was to study the main aspects of integrated end-to-end speech recognition and the use of modern technologies in the natural processing of agglutinative languages, including Kazakh. In this article, the study of language models was carried out using comparative, graphic, statistical and analytical-synthetic methods, which were used in combination. This paper addresses automatic speech recognition (ASR) in agglutinative languages, particularly Kazakh, through a unified neural network model that integrates both acoustic and language modeling. Employing advanced techniques like connectionist temporal classification and attention mechanisms, the study focuses on effective speech-to-text transcription for languages with complex morphologies. Transfer learning from high-resource languages helps mitigate data scarcity in languages such as Kazakh, Kyrgyz, Uzbek, Turkish, and Azerbaijani. The research assesses model performance, underscores ASR challenges, and proposes advancements for these languages. It includes a comparative analysis of phonetic and word-formation features in agglutinative Turkic languages, using statistical data. The findings aid further research in linguistics and technology for enhancing speech recognition and synthesis, contributing to voice identification and automation processes.

Investigating Patient Satisfaction Through Web-Based Reviews of Norwegian Dentists: Quantitative Study Using the Meaning Extraction Method.

Challenging encounters in health care professions, including in dentistry, are relatively common. Challenging encounters can be defined as stressful or emotional situations involving patients that could impact both treatment outcomes and patients' experiences. Through written web-based reviews, patients can share their experiences with health care providers, and these posts can be a useful source for investigating patient satisfaction and their experiences of challenging encounters. This study aims to identify dominant themes from patient-written, web-based reviews of dentists and investigate how these themes are related to patient satisfaction with dental treatment. The study data consisted of 11,764 reviews written by dental patients, which included 1- to 5-star ratings on overall satisfaction and free-text comments. The free-text comments were analyzed using Linguistic Inquiry and Word Count software, and the meaning extraction method was used to group words into thematic categories. These themes were used as variables in a multilevel logistic regression analysis to predict patient satisfaction. Eight themes emerged from the analyses, of which 6 (75%)-explanation (odds ratio [OR] 2.56, 95% CI 2.16-3.04; P<.001), assurance (OR 3.61, 95% CI 2.57-5.06; P<.001), performance assessment (OR 2.17, 95% CI 1.84-2.55; P<.001), professional advice (OR 1.81, 95% CI 1.55-2.13; P<.001), facilities (OR 1.78, 95% CI 1.08-2.91; P=.02), and recommendation (OR 1.31, 95% CI 1.12-1.53; P<.001)-increased the odds of high patient satisfaction. The remaining themes (2/8, 25%)-consequences of treatment need (OR 0.24, 95% CI 0.20-0.29; P<.001) and patient-centered care (OR 0.62, 95% CI 0.52-0.74; P<.001)-reduced the odds of high patient satisfaction. The meaning extraction method is an interesting approach to explore patients' written accounts of encounters with dental health professionals. The experiences described by patients provide insight into key elements related to patient satisfaction that can be used in the education of dental health professionals and to improve the provision of dental health services.

Do pictures' emotional valence and arousal affect younger and older adults' narratives?

Emotional content, specifically negative valence, can differentially influence speech production in younger and older adults' autobiographical narratives, which have been interpreted as reflecting age differences in emotion regulation. However, age differences in emotional reactivity are another possible explanation, as younger and older adults frequently differ in their affective responses to negative and positive pictures. The present experiment investigated whether a picture's valence (pleasantness) and arousal (intensity) influenced older adults' production of narratives about those pictures. Thirty younger and 30 older participants produced narratives about pictures that varied in valence (positive, negative, and neutral) and arousal (high, low). Narratives were recorded via Zoom, transcribed, and analyzed with Linguistic Inquiry and Word Count-22 to get measures of emotional word use, disfluencies, and linguistic distance. Results showed that negative valence increased age differences in speech production independent of picture arousal: Relative to younger adults, older adults used more positive words, fewer negative words, and had more silent pauses when telling narratives about negative pictures. In contrast, high arousal decreased age differences such that older adults used fewer positive words in narratives about positive pictures and more linguistically distant words evidenced by fewer present-tense verbs, relative to narratives about low-arousal pictures. Contrary to an explanation of enhanced regulation or control over emotions in older adulthood, these findings support the idea that older adults' speech production is influenced by their reactivity or affective response to emotional stimuli even when the task is not to communicate one's emotions. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

She Speaks for Millions: The Emergence of Female Diplomatic Voices in the Russo-Ukrainian War

This research critically investigates the public diplomacy strategies deployed by a cohort of influential female European leaders on Twitter during the Russo-Ukrainian War of 2022-2023. The study comprises eight leaders - Kallas (Estonia), Marin (Finland), von der Leyen (President of the European Commission), Metsola (President of the European Parliament), Sandu (Moldova), Simonyte (Lithuania), Zourabichvili (Georgia), and Meloni (Italy) - representing millions of constituents. By mirroring the analytical attention given to Ukraine's President, Volodymyr Zelenskyy, this study scrutinizes the distinct approaches and differences in emotional, cognitive, and structural language use between these influential female figures and President Zelenskyy in their respective Twitter discourse during the crisis. By employing the Linguistic Inquiry and Word Count (LIWC)-22 dictionary, this research intends to uncover unique patterns and nuances in the language use of these leaders, thereby offering fresh insights into the gendered communication and diplomacy employed by the female cohort during conflicts. Keywords: Public diplomacy, Female leadership, Russo-Ukrainian War, Twitter communications, Linguistic Inquiry and Word Count (LIWC), Diplomacy, gendered communication

Elicited imitation in second language acquisition research: New insights to advance methodological rigor (Introduction to the special issue)

This introduction provides an overview of the state of research on and knowledge about the use of elicited imitation in applied linguistics research, highlighting gaps in our current understanding of this tool. It then introduces the eight articles that comprise the special issue.

Specificity and generality of lexical bundles in the rhetorical moves of Applied Linguistics research article introductions

Investigating the linguistic realizations of rhetorical moves through lexical bundles is a research area that has recently gained momentum. However, the lack of a fully annotated corpus has impeded the full understanding of the relationship between formulaic sequences and their corresponding moves. Accordingly, this study addressed this lacuna by annotating 1000 Applied Linguistics research article (RA) introductions using a framework developed for the current study to identify bundles characterizing communicative moves that configure introductions. Following a manual move identification of RA introductions, lexical bundles were extracted using AntConc and manually refined. This was followed by systematic identification of their functions in their co-text. Results indicated that some lexical bundles (i.e., move-general) occur in several moves or steps, whereas other bundles (i.e., move-specific) perform distinct discursive functions within each move. In other words, the development of rhetorical moves and bundles within each move is intricately interwoven and collaboratively contributes to developing ideas in the analyzed section. Furthermore, the functional analyses indicated that text-oriented bundles emerged as a strong indicator of the communicative moves of RA introductions, which performed distinct functions within each move. Drawing on key pedagogical insights, findings can serve as an instructional framework for instructors, students, and novice writers.