Volume 11, Issue 2 -

In Old English, the ditransitive construction with an accusative (direct) object and a dative (indirect) object occurred with two alternating object orders: ACC-DAT vs. DAT-ACC. This study examines the motivations behind the OE speakers’ choice for one of both orders. The effect of 16 factors was evaluated based on a corpus sample of N = 2409 sentences drawn from the York-Toronto-Helsinki Parsed Corpus of Old English Prose (Taylor et al. 2003). The data was analysed by means of a mixed-effects logistic regression analysis. The results indicate that the ACC+DAT alternation was largely driven by the same factors that motivate the dative alternation in later stages of British English. However, no evidence was found for specific verb preferences in Old English, which suggests that the OE object alternation was less driven by semantics than the dative alternation in PDE. It is argued that the results further substantiate Wolk et al.’s (2012) claim that the cognitive mechanisms underlying present-day probabilistic patterns also underlie past variation.

The view that suppression of a concept within the scope of negation is not unconditional was originally introduced by Giora (2003, 2006; Giora and Fein 1999; Giora et al. 2007) via the retention hypothesis . Giora and her colleagues argue that negation does not necessarily suppress the concept within its scope. Instead, it often retains it for pragmatic considerations, both in the mind of the speaker and the addressee. The present study provides a quantitative corpus-based test for the retention hypothesis, that is the non-obligatoriness of suppression of negated concepts (also known as the negation as mitigation hypothesis, Giora 2003; Giora et al. 2005b), via a two-pronged method which combines corpus data and behavioral data. It focuses on the notion of polarity strength , which is a numerical value disclosing the degree of positivity or negativity associated with an adjective. A simple statistic which is introduced for the sake of this study – the Strength Index ( SI ) – naïvely assumes that canonical adjectives can be mitigated by replacing them with their negated antonyms, thus making it possible to attribute SI to them. SI is calculated for 8 canonical adjectival antonymous pairs of an emotive nature (such as good-bad ). Depending on prior positive expectations, the retention hypothesis will gain support if the following results are obtained: Correlation between the SIs of unfavorable adjectives (e.g., bad ) and behavioral data on the one hand, and the lack of correlation between the SIs of favorable adjectives (e.g., good ) and behavioral data, on the other hand. Results attest to this correlation pattern, providing support for the retention hypothesis (see also Colston 1999).

Many applied and corpus linguists entertain the idea of collocations, and other types of formulaic language, being processed as unanalysed , or holistic units. It has, indeed, been demonstrated that, due to their frequency and predictability, formulaic sequences are processed quantitatively faster than matched novel phrases. This finding implies an important role of phrasal frequency in language processing and highlights the contribution of entrenchment of a particular phrasal configuration in memory. This finding, however, cannot be taken to suggest that formulaic sequences are necessarily processed as unanalysed, or holistic units. The present paper reviews some of the recent studies and explains why a processing advantage observed for formulaic sequences over novel phrases should not be equated with holistic storage and processing. The present paper is not intended as an overview of the studies on on-line processing of formulaic language. For a comprehensive review of the method and findings specific to formulaic sequences, their on-line representation and processing, we direct an interested reader to Siyanova-Chanturia (2013) and Siyanova-Chanturia and Martinez (2014).

This article presents a discourse annotation methodology based on Rhetorical Structure Theory and an empirical study of annotating a corpus of specialized medical texts in Basque. The annotation process includes two phases: segmentation and annotation of rhetorical relations. Phase one entails an initial study which leads to establishing linguistic criteria for sentence-based segmentation; a second phase focuses on annotation of rhetorical relations. After establishing discourse segments and rhetorical relations, the annotation process is analyzed and evaluated by means of the method commonly used in RST (Marcu 2000). Inconsistencies detected in the evaluation method lead the authors to redefine some criteria of the evaluation method. As a result of this work, a small annotated Basque-language corpus is provided to scientific community.

Research on Language Technologies demands a wide range of linguistic resources. C-ORAL-JAPON is a corpus of spontaneous spoken Japanese. It has more than 12 recording hours and 149,380 words, divided into three types of texts: monologues, dialogues and conversations. The corpus is made up of 39 text files (transcription) aligned with the corresponding audio files. In total, 59 Japanese native speakers of different ages have participated in the corpus recordings. Thanks to its characteristics, this corpus is an ideal resource for linguistic or sociolinguistic studies, or to be used as working material in teaching Japanese as a foreign language.