The precedent phenomenon is estimated as a productive source for creating the language game in the headlines of political media discourse. Headlines based on a language game draw the attention of the reader more quickly. There are often used precedent phenomena, understood as culturally loaded signs, known to a major part of the representatives of the same national community. That is a precedent phenomenon that may serve multiple purposes. We have focused on its effect applied in the headings. The coincidence of the background knowledge of the author and the reader contributes to the hidden influence on the images of politicians formed in the mind of the reader.
The aim of this paper is to overview the presence of catastrophes in game worlds and, in particular, to investigate what they can tell us about real catastrophes. To this end, we present a semiotic typol-ogy of catastrophes, confronting them with epistrophes and apostrophes and further articulating them relative and absolute cessation events. Then we highlight the long-standing relationship between playfulness and disasters in literature, cinema and video games underlining how the suppos-edly opposite characteristics of the two are, in fact, a very productive cultural trope. To conclude, we look into some examples of catastrophes in game worlds, both relative (such as the “corrupted blood incident” in World of Warcraft) and absolute (the end of the worlds in StarWars Galaxies and Matrix Online).
The article examines the individual and social, practical, and theoretical presumptions (“idols” and “beliefs”) that constitute the conscious and unconscious re-construction of the social reality and reality of different conventional sign systems that represent and are represented by society. It is shown that in everyday life and in theoretical studies, we quite often analyze sign systems as if they were autonomous and empirically “given” realities. The work explains how this “natural belief” originated and developed. It is argued that conventional sign systems cannot be reduced to the reality of material “sign vehicles” because in society, sign systems are both subjective and objective, internal and external, and process and object.
The Hittite grammar is characterized by a morphosyntactic split that affects the behaviour of the inflectional classes of Noun phrases (DPs). While a singular neuter transitive subject is marked by /-anza/suffix, commons DPs end with an /-š/mark. In addition, intransitive neuter subjects and neuter objects pattern in the same way, marked by /-ø/, while in commons the object role is marked by an /-n/ ending, which distinguishes it from the subjects. The aim of this paper is to investigate over a possible definition of split ergativity in the Hittite grammar.
Kateřina Rysová, Magdaléna Rysová, Michal Novák, Jiří Mírovský and Eva Hajičová
In the paper, we present EVALD applications (Evaluator of Discourse) for automated essay scoring. EVALD is the first tool of this type for Czech. It evaluates texts written by both native and non-native speakers of Czech. We describe first the history and the present in the automatic essay scoring, which is illustrated by examples of systems for other languages, mainly for English. Then we focus on the methodology of creating the EVALD applications and describe datasets used for testing as well as supervised training that EVALD builds on. Furthermore, we analyze in detail a sample of newly acquired language data – texts written by non-native speakers reaching the threshold level of the Czech language acquisition required e.g. for the permanent residence in the Czech Republic – and we focus on linguistic differences between the available text levels. We present the feature set used by EVALD and – based on the analysis – we extend it with new spelling features. Finally, we evaluate the overall performance of various variants of EVALD and provide the analysis of collected results.
Daniel Kondratyuk, Ronald Cardenas and Ondřej Bojar
Recent developments in machine translation experiment with the idea that a model can improve the translation quality by performing multiple tasks, e.g., translating from source to target and also labeling each source word with syntactic information. The intuition is that the network would generalize knowledge over the multiple tasks, improving the translation performance, especially in low resource conditions. We devised an experiment that casts doubt on this intuition. We perform similar experiments in both multi-decoder and interleaving setups that label each target word either with a syntactic tag or a completely random tag. Surprisingly, we show that the model performs nearly as well on uncorrelated random tags as on true syntactic tags. We hint some possible explanations of this behavior.
The main message from our article is that experimental results with deep neural networks should always be complemented with trivial baselines to document that the observed gain is not due to some unrelated properties of the system or training effects. True confidence in where the gains come from will probably remain problematic anyway.
Implicit causality of interpersonal transitive verbs (IC) pertains to preferences to attribute the cause of a given action to the subject or the object referent in active clauses. Causal attribution is operationalized as the probability of referential continuation in a subsequent explanatory clause. This paper presents an explorative investigation into the causal biases of action verbs, which in contrast to affective verbs have received less attention in IC research. We approach implicit causality as a discourse level phenomenon based on the textual level of discourse representation and enriched by conceptual knowledge. In study 1, we targeted IC effects of German action verbs (N = 52) in sentences containing causal, additive and adversative connectives. Results showed that IC based categories of subject-object-, and non-biasing predicates were clearly discernable in causal contexts only. In study 2, we examined effects of situational knowledge (physical affectedness & social acceptability) and affective appraisals (valence & arousal) represented in the conceptual structure of the verbs on the construal of causality biases and their interplay with immediate contextual information such as gender of referents. Results show that higher degrees of physical affectedness were associated with causal attribution to the object referent. This effect was modulated by the affective properties of the verbs. Our findings revealed the influence of physiological arousal, an affective dimension not considered in previous investigations of IC. Actions with a strong physical impact that were characterized by high arousal, e.g., kick, or tickle were more likely to be explained with reference to the subject. Participants also considered the available contextual information, as indicated by the significant interactions of gender information with arousal. Within the subsample of non-biasing verbs, higher estimates for social behavior increased probabilities of causal attributions to the subject.
Singular nouns in the scope of a distributive operator have been shown to be treated as conceptually plural (Patson and Warren, 2010). The source of this conceptual plurality is not fully clear. In particular, it is not known whether the concept of plurality associated with a singular noun originates from distributing over multiple objects or multiple events. In the present experiment, iterative expressions (distribution over events) were contrasted with collective and distributive sentences using a Stroop-like interference technique (Berent, Pinker, Tzelgov, Bibi, and Goldfarb, 2005; Patson and Warren, 2010). A trend in the data suggests that event distributivity does not elicit a plural interpretation of a grammatically singular noun, however the results were not statistically significant. Possible causes of the non-significant results are discussed.