Proximity hype as a frame attribute in the reporting on North Korea: Frame strength in competitive contexts

This study analyses generic frames exhibiting varying strength when in a competitive context in an effort to identify the most recurring and repeated patterns and schemata in the framing of global news. The two-level analysis revealed six recurring attributes (official and/or credible sources, repetition, journalistic lexical bias, proximity hype, episodic nature of frame and negativity bias) that are believed to encompass frames and influence their power of persuasiveness. Special focus is placed on proximity hype with its three dominant angles recurring in the publications, which implies a common use of this aspect in the framing of news content.

Machine learning for learner English

This paper discusses machine learning techniques for the prediction of Common European Framework of Reference (CEFR) levels in a learner corpus. We summarise the CAp 2018 Machine Learning (ML) competition, a classification task of the six CEFR levels, which map linguistic competence in a foreign language onto six reference levels. The goal of this competition was to produce a machine learning system to predict learners’ competence levels from written productions comprising between 20 and 300 words and a set of characteristics computed for each text extracted from the French component of the EFCAMDAT data (Geertzen et al., 2013). Together with the description of the competition, we provide an analysis of the results and methods proposed by the participants and discuss the benefits of this kind of competition for the learner corpus research (LCR) community. The main findings address the methods used and lexical bias introduced by the task.

The Lexical Bias Effect during Speech Production in the First and Second Language

The lexical bias effect is the tendency for people to make phonological speech errors that result in existing words. Several studies have argued that this effect arises from a combination of factors: the self-monitoring system covertly weeding out more nonword than word errors and feedback of activation during speech production biasing towards lexical outcomes. Moreover, lexicality of the context has been shown to influence the occurrence of the lexical bias effect (Hartsuiker, Corley, & Martensen, 2005), supporting a role for monitoring. But how does this process differ in one’s first language (L1) as opposed to this same process in the second language (L2) and is there even a difference to begin with? To address that question, we tested whether people also show the lexical bias effect when speaking in a second language (L2) and if so, whether the effect is also modulated by context lexicality. Additionally, we tested whether recent exposure to existing words in L2 influences such a lexical bias effect. We observed a lexical bias effect in L1 but not in L2 in Experiment 1. Moreover, the lexical bias effect in L1 was marginally modulated by context. Experiment 2, in which more L2 words were presented, did not demonstrate a lexical bias effect in either language. However, an analysis of a subset of the data (namely the blocks that were identical in both experiments and thus directly comparable) found a three-way interaction between Outcome, Language, and Experiment. This interaction suggests a strong lexical bias effect in Experiment 1 for L1 but not for L2 whereas Experiment 2 reveals a comparable lexical bias effect in both languages. This indicates that more exposure to existing L2 words leads to an increase in activation of the lexical representations of the target language, thereby increasing the number of transpositions and therefore an increase in the strength of the lexical bias effect.