Corpus-based Methodology Research Articles

A lectometric analysis of aggregated lexical variation in written Standard English with Semantic Vector Space models

Lectometry is a corpus-based methodology that explores how multiple language-external dimensions shape language usage in an aggregate perspective. The paper combines this methodology with Semantic Vector Space modeling to investigate lexical variability in written Standard English, as sampled in the original Brown family of corpora (Brown, LOB, Frown and F-LOB). Based on a joint analysis of 303 lexical variables, which are semi-automatically extracted by means of a SVS, we find that lexical variation in the Brown family is systematically related to three lectal dimensions: discourse type (informative versus imaginative), standard variety (British English versus American English), and time period (1960s versus 1990s). It turns out that most lexical variables are sensitive to at least one of these three language-external dimensions, yet not every dimension has dedicated lexical variables: in particular, distinctive lexical variables for the real time dimension fail to emerge.

Evidential adverbials in Lithuanian: a corpus-based study

The present study examines the functional distribution of the adverbials akivaizdžiai ‘evidently’, aiškiai ‘clearly’, ryškiai ‘visibly/clearly’, tariamai ‘allegedly/supposedly’ and aišku ‘clearly/of course’ in Lithuanian fiction and academic discourse. The aim of the study is to identify the evidential and/or pragmatic functions of perception and communication-based adverbials which can be traced synchronically to different syntactic environment (a predication manner adverbial and a CTP clause). The paper examines the frequency of these adverbials, their position, scope, functions, co-occurrence with argumentative markers, word class (adverb or non-agreeing adjective) and the type of discourse they occur in. The research is conducted by applying a corpus-based methodology and the data are obtained from the Corpus of the Contemporary Lithuanian Language, namely from the subcorpus of fiction, and the Corpus of Academic Lithuanian. The perception-based adverbials akivaizdžiai ‘evidently’, aiškiai ‘clearly’, ryškiai ‘clearly/visibly’ and aišku ‘clearly/of course’ denote inferences drawn from perceptual and conceptual evidence and contribute to persuasive authorial argumentation, while the communication-based adverbial tariamai ‘allegedly/supposedly’ functions as a hearsay marker. The latter may also be used as an epistemic marker which refers to unreal or imagined situations. In contexts of common knowledge, the adverbial aišku ‘clearly/of course’ acquires interactional and textual functions and thus reveals traces of pragmaticalisation. In academic discourse, it signals interaction with the addressee and links units of discourse, while in fiction it functions as a speech act modifier in a variety of emotive contexts. The pragmaticalisation of aišku ‘clearly/of course’ is also marked by its high frequency, positional mobility (initial, medial, final) and scopal variability (clausal, phrasal). Alongside its discrete evidential and pragmatic functions, the adverbial aišku ‘clearly/of course’ displays the merger of the two functions. The adverbials akivaizdžiai ‘evidently’, aiškiai ‘clearly’, ryškiai ‘visibly/clearly’ and tariamai ‘allegedly/supposedly’ do not acquire a pragmatic function, which is indicated by their frequency and position. The results of the present study corroborate the findings of previous studies that common sources of evidential adverbials and pragmatic markers in Lithuanian are verb-based, adjective-based and noun-based CTP clauses.

Phraseological profiles of legislative genres: complex prepositions as a special case of legal phrasemes in EU law and national law

The paper explores the hypothesis that a large proportion of non-terminological word combinations in legislation is built around complex prepositions, which significantly contributeto phraseological profiles of legislative genres. The paper analyses the distribution and functions of complex prepositions in multilingual EU law and national law, on a comparative (cross-systemic) and contrastive English-Polish basis, against the background of general language. The analysis is conducted in the corpus-based methodology with the corpora of EU legislation (JRC Acquis) – regulations and directives, national legislation of the UK (BoLC) and of Poland (PLC), and general corpora (BNC and NKJP). The findings confirm that complex prepositions are very frequent and hence cognitively salient in the genre of legislation: complex prepositions show increased distribution against general language, in particular in Polish. It is demonstrated that national legislation and EU legislation (translationese) are profiled by different sets of salient prepositions, which may adversely affect the readability of the latter due to interference. Functionally, it has been demonstrated that the phraseological profiles of legislative instruments are marked by complex prepositions used predominantly in referencing patterns (authority, conflict), conditionals, anchoring (framing) patterns, defining patterns and time deixis.

The shifting representation of common people in China’s news media

This paper examines the meanings associated with the high-frequency social address term baixing (‘common people’) in China’s contemporary news media, through the analysis of a sizable corpus of data from China’s online news portals, together with the results of qualitative research derived from participation in an online discussion forum. The findings of this project indicate that the term baixing is closely associated with discourses of insignificance, obedience, and powerlessness. This study sets out to make a contribution to current research on Chinese address terms through the use of a corpus-based methodology, by expanding the site of investigation beyond face-to-face encounters to a wider media public, and thereby to provide insights into the discourses of citizen identity in contemporary China.

Language, corpus and empowerment: applications to deaf education, healthcare and online discourses

Language, Corpus and Empowerment applies a novel corpus-driven approach to the exploration of the concept of empowerment in healthcare. The book proposes an innovative corpus-based methodology for finding evidence of empowerment in language use, using data from a video intervention delivered to families of deaf children, as well as assessing the effects of the intervention on the family. Language, Corpus and Empowerment provides a working definition of empowerment which incorporates concepts from linguistics and learning theory; uses corpus analysis to provide evidence of how video interventions can transform peoples perspectives; examines this new methodology as a potential tool for analysing conversational data longitudinally and at a case-by-case level; demonstrates how a corpus-based methodological approach can be applied in conjunction with other language-based approaches, such as discourse analysis and conversation analysis, to explore the ways in which complex social processes occur in interaction; makes a valuable development in the assessment of the impact of healthcare interventions and the language of empowerment. Insightful and ground-breaking, Language, Corpus and Empowerment is essential reading for anyone undertaking research within corpus linguistics.

A Corpus-based investigation of critical reflective practice and context in early career teacher settings

Reflective practice is at the core of teacher education programmes and is highly regarded as an essential component in the education of new and experienced teachers. Given the recent interest in language use and the role of discourse in articulating knowledge of one’s practice, this paper focuses on how two groups of early career teachers from distinct teaching contexts engage in reflection. The study aims to (i) investigate how they reflect and (ii) explore to what extent their reflections may be influenced by their contexts. The paper mines two small written sub-corpora, which are part of the larger 200,000-word Corpus of Reflective Practice (CoRP). Using corpus-based tools and methodologies, the paper identifies distinct trends which reveal insights into different conceptualisations of reflective practice. As a result, the paper raises awareness of the need to consider (inter)national cultural sensitivity, and questions the assumption that early career reflective practice is flexible enough to work across cultural, social and institutional differences and influences without more careful guidance and consideration. The paper also highlights a predominant shared trait across the data which signals the need to see reflective practice as a process and to value the kind of reflective practice in which early practitioners engage as being reflective of their stage in the process.

Estudos de Terminologia no Brasil: diálogos com Portugal

This paper contrasts some texts that deal with the history of the terminology research in Brazil, especially the research aimed to extract or recognize terminology in corpora, a widespread practice among us only from the year 2000, with pioneering texts on this topic produced by Portuguese researchers, represented here by Belinda Maia. The intention is to recognize her role as disseminator of the corpus-based methodologies. The paper follows showing how the dialogue between the Terminology studies from Brazil and Portugal is important for the promotion of Portuguese language in the global scenario of scientific and technical communication.

Centre for Applied Language Studies, University of Limerick, Ireland

The Centre for Applied Language Studies (CALS), founded in 1997, brings together researchers and postgraduate students from several disciplines in language studies, and is structured in three research clusters: New learning environments; Discourse, society and identity; and Plurilingualism and language policy. There is a certain amount of overlapping between the clusters, and several researchers are active in more than one cluster. Thus research in language teacher education is present both in the New learning environments cluster and in Discourse, society and identity. Corpus-based methodologies are also prominent in both these clusters. Furthermore, discourse analysis is present as a methodology in all three. The following sections provide information on current research and selected research outcomes within the three research clusters from 2011 to 2013.

Leaders and Leadership: A Corpus-Based Approach for Evaluating the Evolution of Leadership

Much has been written about leaders and leadership, managers and management. The terms are often used interchangeably. (see http://thesaurus.reference.com). With rapid changes in technology, and the spread of capitalism, networks which use English as the common language are flourishing. Many authors have written about the affects this has had on management and leadership and the differences between the two. This paper reviews common definitions of leadership, explores some of the key concepts associated with the transition from the industrial to the information age, and the evolution of the terminology of leadership, and then reviews the results of a preliminary research project which uses a corpus-based methodology to analyse the text in a diachronic set of documents collated from the Leadership Quarterly 1990 to 2005. Future research will build on this study as the basis for reviewing other sources and domains in the field of leadership, modelling the relationship between cross-cultural issues and leaders in organizations.

A New Corpus-Based Methodology for Pedagogical Vocabulary Selection : Compilation of "HEV1800" for Japanese High School Students(Theoretical Studies and Experimental Studies,The 44th CELES Conference in Yamanashi)

A New Corpus-Based Methodology for Pedagogical Vocabulary Selection : Compilation of "HEV1800" for Japanese High School Students(Theoretical Studies and Experimental Studies,The 44th CELES Conference in Yamanashi)

Covering Muslim women: Semantic macrostructures in BBC News

Despite a proliferation of research on Islam and Muslims in the media, very little work has focused on Muslim women, a much-debated social group that merits special consideration. This article aims to investigate how Muslim women are represented in BBC News website texts using a purpose-built corpus (1.9 million words in 3269 articles). The research employs analytical tools from the discourse-historical, socio-cognitive, and sociosemantic approaches to critical discourse studies. These are combined with corpus-based methodologies to investigate the semantic macrostructures that tend to be associated with Muslim women and the discursive strategies employed in the representation of the hijab. The study is novel in its exhaustive approach to identifying salient and underreported issues related to Muslim women in news discourse. It also introduces a more integrated analysis that combines manual and automated techniques and tackles the quantification of qualitative results where possible. Findings suggest that Muslim women’s representations are largely restricted in terms of regional coverage. Semantic macrostructures related to conflict and crime are prevalent. The hijab remains a nodal discourse surrounding Muslim women whose function as a descriptive feature is often unclear, raising serious questions about its relevance. A number of recommendations are made for journalists to become cognizant of their own context models when reporting on Muslim women.

Shame, embarrassment and guilt: Corpus evidence for the cross-cultural structure of social emotions

AbstractThis article is a cross-cultural and cross-linguistic comparison of three interrelated emotional categories of shame, embarrassment and guilt in two different cultural settings of individualistic societies, as represented here by Britain and America, and a collectivist society, such as Poland. The conceptual field of SHAME is operationalized through its three near-synonymous adjectival exponents, "ashamed"/"zawstydzony", "embarrassed"/"zażenowany", and "guilty"/"winny". Drawing on relevant research in social and cognitive psychology as well as linguistics, the present study applies advanced quantitative corpus-based methodology to reconstruct the cultural and conceptual profiles of the three emotions.

Semantic similarity measures in the biomedical domain by leveraging a web search engine.

Various researches in web related semantic similarity measures have been deployed. However, measuring semantic similarity between two terms remains a challenging task. The traditional ontology-based methodologies have a limitation that both concepts must be resided in the same ontology tree(s). Unfortunately, in practice, the assumption is not always applicable. On the other hand, if the corpus is sufficiently adequate, the corpus-based methodologies can overcome the limitation. Now, the web is a continuous and enormous growth corpus. Therefore, a method of estimating semantic similarity is proposed via exploiting the page counts of two biomedical concepts returned by Google AJAX web search engine. The features are extracted as the co-occurrence patterns of two given terms P and Q, by querying P, Q, as well as P AND Q, and the web search hit counts of the defined lexico-syntactic patterns. These similarity scores of different patterns are evaluated, by adapting support vector machines for classification, to leverage the robustness of semantic similarity measures. Experimental results validating against two datasets: dataset 1 provided by A. Hliaoutakis; dataset 2 provided by T. Pedersen, are presented and discussed. In dataset 1, the proposed approach achieves the best correlation coefficient (0.802) under SNOMED-CT. In dataset 2, the proposed method obtains the best correlation coefficient (SNOMED-CT: 0.705; MeSH: 0.723) with physician scores comparing with measures of other methods. However, the correlation coefficients (SNOMED-CT: 0.496; MeSH: 0.539) with coder scores received opposite outcomes. In conclusion, the semantic similarity findings of the proposed method are close to those of physicians' ratings. Furthermore, the study provides a cornerstone investigation for extracting fully relevant information from digitizing, free-text medical records in the National Taiwan University Hospital database.

A contrastive analysis of (English) ‘there’ and (Spanish) hay existential sentences

This study is devoted to the grammatical, semantic and informative analysis of the so-called existential sentence (“There is a girl in the garden” in English, or Hay una niña en el jardín ‘There-is a girl in the garden’ in Spanish) in an attempt to establish a multi-linguistic prototype of the construction. To that end, data from several corpora of contemporary spoken English and Spanish are analysed in a number of ways, including the frequency of this construction in the two languages, the basic elements of its syntactic structure, and the semantic and informative constraints which operate in the existential/presentational construction. This study also deals with the degree of variation which these sentences exhibit and how this affects the selection of the marker of the construction (‘there’, hay), agreement between the marker or the verb and the postverbal noun phrase, the accommodation of additional constituents such as locative phrases or nominal postmodifiers and complements, the so-called indefiniteness restriction, and the compliance with general informative principles to which English and Spanish are claimed to be subject. A corpus-based contrastive methodology leads both to a prototypical and to a language-specific description of the existential construction in English and Spanish, in which the notion of grammatical, semantic and informative versatility plays a significant role.

Sharing and collaborating between an online community of novice teachers: CMC in language teacher education

New and emerging technologies have transformed the classroom (Mishra and Koehler, 2006) and continue to do so, and it has been reported that teachers and teacher educators now work in ever evolving environments (Elliott, 2009). Research indicates that varied forms of CMC can be implemented to foster collaborative and social learning (Arnold and Ducate, 2006), and the formation of communities of practice (CoPs) (Arnold et al., 2005). Therefore, this paper investigates the implementation of online and face-to-face (F2F) communication in an English Language Teaching (ELT) teacher education programme, with the aim of illuminating the potential such modes have for sharing and collaborating, for providing a space for situated learning and open communication. Results are analysed using a corpus-based methodology, drawing on three aspects of community membership, namely a joint enterprise, mutual engagement and a shared repertoire (Wenger, 1998). This paper closes with a discussion of the implications of such data for Language Teacher Education (LTE) in a technologically-oriented world.

More spoken or more translated?

Since the early 1990s, with the advance of computerized corpora, translation scholars have been using corpus-based methodologies to look into the possible existence of overriding patterns (tentatively described as universals or as laws) in translated texts. The application of such methodologies to interpreted texts has been much slower in developing than in the case of translated ones, but significant progress has been made in recent years. After presenting the fundamental methodological hurdles—and advantages—of working on machine-readable (transcribed) oral corpora, we present and discuss several recent studies using cross-modal comparisons, and examine the viability of using interpreted outputs to explore the features that set simultaneous interpreting apart from other forms of translation. We then set out to test the hypothesis that modality may exert a stronger effect than ontology—i.e. that being oral (vs. written) is a more powerful influence than being translated (vs. original).

Productivity of verb-forming suffixes in Modern Greek: a corpus-based study

The paper studies the productivity of eight Greek verb-forming suffixes (-aro, -eno, -evo, -ino, -izo, -(i)azo, -jazo and -ono) in a corpus of 4,143,583 words. The corpus used comes from the Corpus of Greek Texts and includes all text types found in it in the same proportion. Our study follows similar research on other languages, investigating frequency and morphological productivity with a corpus-based methodology. The productivity of the suffixes was studied in the corpus as a whole, across its two subcorpora of written and spoken registers and across and within the text types included in the corpus. Three main sets of suffixes were identified: (a) the very productive -izo and -ono, (b) the moderately productive -evo and -aro and (c) the least productive or unproductive -iazo, -jazo, -eno and -ino. The results from the study of text types suggest that the degree of productivity of each suffix is influenced by the text type in which it occurs. Quantitative evidence drawn from the measurement of productivity, by means of vocabulary growth curves and related methods, was combined with qualitative evidence on the semantic and other properties of verb-forming suffixes. On the basis of our results it cannot be argued that there is a default suffix for verb derivation in Modern Greek; it is rather suggested that the two most productive suffixes, -izo and -ono, are used in Greek in a more or less complementary way, since they show a different preference for meanings and text types.

Contextualisation in East African English: A corpus-based study of register variation

In the article, contextualisation refers to the anchoring of discourse in a specific time and place. Five linguistic features associated with spatio-temporal contextualisation were analysed: present tense verbs, past tense verbs, perfect aspect, time adverbials and place adverbials. The purpose of the research was to identify whether these linguistic features are used equally across the board in East African English discourse, or whether certain features are associated with particular registers more often. Twentytwo spoken and written registers of the International Corpus of East African English (ICE-EA) were analysed using a corpus-based methodology. Results indicate that on the one hand, social letters have the highest standardised score for the contextualisation features of all the registers. On the other hand, instructional writing such as administrative texts has the lowest standardised score for these features. Overall, the different needs and uses for contextualisation are reflected in the total standardised scores across registers.

Translator-oriented, corpus-driven technical glossaries: the case of cooking terms

Although there are many English–Portuguese technical glossaries on the market, very few are designed to meet the specific needs of translators, whose main task is the production of an idiomatic text either in the mother tongue or in the foreign language. For that purpose, a mere list of monolexical terms (usually based on previous compilations or abstract conceptual schemata, and their equivalents) will not suffice. Most importantly, the producer of a text needs to know how a word is used in context, and this can be inferred from the observation of authentic examples. Corpus Linguistics has proved to be an invaluable tool in retrieving technical terms and phraseologies from corpora. In this paper, we employ a corpus-based methodology or, more precisely, the ‘corpus-driven approach’, to compile a bilingual, monodirectional English–Portuguese glossary of cooking terms ( Teixeira and Tagnin, 2008 ), covering the various stages of the project, with an emphasis on the identification of candidate terms, and their subsequent validation through the generation of frequency and keyword lists using the lexical analysis software, WordSmith Tools.

Translator-oriented, corpus-driven technical glossaries: the case of cooking terms

Although there are many English–Portuguese technical glossaries on the market, very few are designed to meet the specific needs of translators, whose main task is the production of an idiomatic text either in the mother tongue or in the foreign language. For that purpose, a mere list of monolexical terms (usually based on previous compilations or abstract conceptual schemata, and their equivalents) will not suffice. Most importantly, the producer of a text needs to know how a word is used in context, and this can be inferred from the observation of authentic examples. Corpus Linguistics has proved to be an invaluable tool in retrieving technical terms and phraseologies from corpora. In this paper, we employ a corpus-based methodology or, more precisely, the ‘corpus-driven approach’, to compile a bilingual, monodirectional English–Portuguese glossary of cooking terms (Teixeira and Tagnin, 2008), covering the various stages of the project, with an emphasis on the identification of candidate terms, and their subsequent validation through the generation of frequency and keyword lists using the lexical analysis software, WordSmith Tools.