Crisis and Language in Ray Bradbury’s The Last Night of the World

Language use in social crisis situations is usually described as being highly ideological, and it exhibits features of affect involving the use of negative evaluation of the perceived social enemies. The present study aims to explore the characters’ language use in Ray Bradbury’s short story entitled The Last Night of the World from a pragma-stylistic perspective. The fictional dialogue that takes place between the two protagonists creates and reflects the dynamics between them, where the unspeakable is only inferred rather than communicated. The analysis reveals special features of verbal communication in a crisis situation, especially focusing on the lexical and morphosyntactic properties, as well as on the verbal interaction and cooperation between the characters revealing their alignment. The results of the analysis prove that the verbal and non-verbal communication between the protagonists do not show the features described in crisis communication; therefore, the text of the story can be interpreted as subverting the generic language use in a critical situation.

Text Classification Model Enhanced by Unlabeled Data for LaTeX Formula

Generic language models pretrained on large unspecific domains are currently the foundation of NLP. Labeled data are limited in most model training due to the cost of manual annotation, especially in domains including massive Proper Nouns such as mathematics and biology, where it affects the accuracy and robustness of model prediction. However, directly applying a generic language model on a specific domain does not work well. This paper introduces a BERT-based text classification model enhanced by unlabeled data (UL-BERT) in the LaTeX formula domain. A two-stage Pretraining model based on BERT(TP-BERT) is pretrained by unlabeled data in the LaTeX formula domain. A double-prediction pseudo-labeling (DPP) method is introduced to obtain high confidence pseudo-labels for unlabeled data by self-training. Moreover, a multi-rounds teacher–student model training approach is proposed for UL-BERT model training with few labeled data and more unlabeled data with pseudo-labels. Experiments on the classification of the LaTex formula domain show that the classification accuracies have been significantly improved by UL-BERT where the F1 score has been mostly enhanced by 2.76%, and lower resources are needed in model training. It is concluded that our method may be applicable to other specific domains with enormous unlabeled data and limited labelled data.

When gender matters in scientific communication: The role of generic language

Prior research has documented gender differences in self-presentation and self-promotion. For example, a recent analysis of scientific publications in the biomedical sciences reveals that articles with women in lead author positions (first and last) included fewer positive words to describe their results than articles with men in lead author positions. Here we examined the role of gender in peer-reviewed publications in psychology, with a focus on generic language. When authors describe their results using generic statements (e.g., “Introverts and extraverts require different learning environments”), those statements gloss over variability, frame an idea as broad, timeless, and universally true, and have been judged to be more important. In a sample of 1,149 psychology articles published in 2015-16 from 11 journals, we found that women in lead author positions were less likely to employ generic language than men in lead author positions, and that publications with more generic language received more citations (as did publications authored by men). We discuss how a subtle gender difference in self-presentation may have direct consequences for how a scientific finding is interpreted and cited, with potential downstream consequences for career advancement for women and men.

A Survey on Sentiment Analysis and Opinion Mining in Greek Social Media

As the amount of content that is created on social media is constantly increasing, more and more opinions and sentiments are expressed by people in various subjects. In this respect, sentiment analysis and opinion mining techniques can be valuable for the automatic analysis of huge textual corpora (comments, reviews, tweets etc.). Despite the advances in text mining algorithms, deep learning techniques, and text representation models, the results in such tasks are very good for only a few high-density languages (e.g., English) that possess large training corpora and rich linguistic resources; nevertheless, there is still room for improvement for the other lower-density languages as well. In this direction, the current work employs various language models for representing social media texts and text classifiers in the Greek language, for detecting the polarity of opinions expressed on social media. The experimental results on a related dataset collected by the authors of the current work are promising, since various classifiers based on the language models (naive bayesian, random forests, support vector machines, logistic regression, deep feed-forward neural networks) outperform those of word or sentence-based embeddings (word2vec, GloVe), achieving a classification accuracy of more than 80%. Additionally, a new language model for Greek social media has also been trained on the aforementioned dataset, proving that language models based on domain specific corpora can improve the performance of generic language models by a margin of 2%. Finally, the resulting models are made freely available to the research community.

Generics in society

ABSTRACTThis article examines two interrelated issues: (i) how considering generics within their social contexts of use contributes to theories of generics, and (ii) how contemporary work on generics provides promising directions for the study of language as an aspect of social life. Examining the function of generics in meaningful interactions stands in contrast to standard treatments, which consider generics as isolated, context-free propositions. Additionally, recent psychological approaches suggest new questions that can enrich sociolinguistic and linguistic anthropological research. These include, for example, when and why generics serve not just negative functions (such as stereotyping) but also positive functions (such as meaning-making), how generics gain their power from what is unstated as opposed to stated, and how generic language distorts academic writing. Ultimately, the study of language in society has the potential to enrich the study of generics beyond what has been learned from their study in linguistics, philosophy, and psychology. (Generics, concepts, categories, stereotyping, induction)*

“Master” of None: Institutional Language Change Linked to Reduced Gender Bias

Concern that masculine generic language (e.g., man to mean humanity) perpetuates gender inequity has led several institutions to formally discourage its use. While previous experimental research indicates that generic terms like man bring more exemplars of men than women to mind, only recently have researchers begun exploring additional consequences of gendered language. Understanding the range of processes affected is of particular importance when evaluating real-world policies. Yale University recently changed the title of a leadership role from master to head. The present study (N=341) investigated what exemplars come to mind (i.e., cognitive accessibility) while also probing memory for women and men in the leadership role both before and after Yale’s language policy change. Students exposed to master generated a male exemplar more than would be expected by the incidence of men and recognized actual men than women in the role more accurately (d’) in a face recognition task. Among students exposed to head, both biases were eliminated. The previous literature shows that masculine generic language brings men to mind. The present work demonstrates a similar effect but in an applied context while further documenting consequences for memory. Gender-inclusive language polices have potential to reduce gender biased thinking with applied significance.

The Unintended Consequences of the Things We Say: What Generic Statements Communicate to Children About Unmentioned Categories

Adults frequently use generic language (e.g., “Boys play sports”) to communicate information about social groups to children. Whereas previous research speaks to how children often interpret information about the groups described by generic statements, less is known about what generic claims may implicitly communicate about unmentioned groups (e.g., the possibility that “Boys play sports” implies that girls do not). Study 1 (287 four- to six-year-olds, 56 adults) and Study 2 (84 four- to six-year-olds) found that children as young as 4.5 years draw inferences about unmentioned categories from generic claims (but not matched specific statements)—and that the tendency to make these inferences strengthens with age. Study 3 (181 four- to seven-year-olds, 65 adults) provides evidence that pragmatic reasoning serves as a mechanism underlying these inferences. We conclude by discussing the role that generic language may play in inadvertently communicating social stereotypes to young children.