Papers reviewed and determined not to be word norm studies. Use the flag icon to report errors or suggest re-inclusion.
18625 papers
This paper introduces a new experimental protocol for studying mental representations of urban soundscapes through a simulation process. Subjects are asked to create a full soundscape by means of a dedicated software tool, coupled with a structured sound data set. This paradigm is used to characterize urban sound environment representations by analyzing the sound classes that were used to simulate the auditory scenes. A rating experiment of the soundscape pleasantness using a seven-point bipolar semantic scale is conducted to further refine the analysis of the simulated urban acoustic scenes. Results show that (1) a semantic characterization in terms of presence/absence of sound sources is an effective way to characterize urban soundscape pleasantness, and (2) acoustic pressure levels computed for specific sound sources better characterize the appraisal than the acoustic pressure level computed over the overall soundscape.
The article analyzes the lexical content of the concept of “tolerance” in the materials published on the pages of high-quality Russian press, presented by “Nezavisimaya Gazeta”, “Literaturnaya Gazeta” and “Kommersant” newspaper. The word “tolerance” means “patience” at the level of lexical meaning are contradictory. This is due to the fact that the scientific literature treats tolerance and patience differently – from synonymy to complete opposition. The presence of the word in the media text “encourages” it to take the ideological orientation of the General public context. Because of the arising of the topical connotations of tolerance of the phenomenon may go to tolerance is the word. As a complex interdisciplinary category “tolerance” in periodicals functions in different discourses and with different evaluation characteristics. Its presence on the pages of print media is often supplemented by certain lexical markers that do not affect the connotations: “political correctness”, “multiculturalism”, “European values”, “globalization”. The semantic content of tolerance in media texts is expressed by associative rows and oppositions, while many lexemes can simultaneously be identified with tolerance and opposed to it. These include: “respect”, “understanding”, “goodwill”, “indifference”, “alienation”, “internationalism”. In texts with a negative assessment of tolerance, the semantic content of the concept is supplemented by the following components: indifferentism, the lack of the concept of norm and pathology, the destruction of traditional norms and values, the loss of national identity. This kind of tolerance, including at the conceptual level, in media journalism is opposed to the primordial Russian ideas about the spiritual values of life. The lexical component of the category “tolerance” in the press reflects the peculiarities of its functioning in the modern linguistic picture and socio-political space of Russia. The texts of high-quality print media provide meaningful language material not only for the philological study of the category of “tolerance”, but also for its interdisciplinary development. At the same time, interdisciplinary comments are necessary as a broad (near and far) context and in the immanent study of language material relating to such an urgent problem as tolerance.
Machine translation (MT) is directly linked to its evaluation in order to both compare different MT system outputs and analyse system errors so that they can be addressed and corrected. As a consequence, MT evaluation has become increasingly important and popular in the last decade, leading to the development of MT evaluation metrics aiming at automatically assessing MT output. Most of these metrics use reference translations in order to compare system output, and the most well-known and widely spread work at lexical level. In this study we describe and present a linguistically-motivated metric, VERTa, which aims at using and combining a wide variety of linguistic features at lexical, morphological, syntactic and semantic level. Before designing and developing VERTa a qualitative linguistic analysis of data was performed so as to identify the linguistic phenomena that an MT metric must consider (Comelles et al. 2017). In the present study we introduce VERTa’s design and architecture and we report the experiments performed in order to develop the metric and to check the suitability and interaction of the linguistic information used. The experiments carried out go beyond traditional correlation scores and step towards a more qualitative approach based on linguistic analysis. Finally, in order to check the validity of the metric, an evaluation has been conducted comparing the metric’s performance to that of other well-known state-of-the-art MT metrics.
This paper investigates the lexical immigration of “arabismos” (Arabic loanwords) in the Spanish language. It compares the differences between old arabismos (integrated during the Middle Ages) and modern arabismos (integrated in contemporary time) in terms of their adaptation to the Spanish norms of spelling, phonetics, semantics and morphosyntactic. It also analyses the use of the arabismos and their immigration process during their first phase of presence in the written Spanish language (or predecessor language to Spanish). Through the study of a corpus collected from the databases of the Royal Spanish Academy (RAE), this paper has come to conclude that the anticipated differences between the two types of arabismos has been affirmed, and that the loss of adaption of the modern arabismos to the Spanish linguistic patterns has resulted in a maintenance of them as foreignisms.
Literary novels are said to distinguish themselves from other novels through conventions associated with literariness. We investigate the task of predicting the literariness of novels as perceived by readers, based on a large reader survey of contemporary Dutch novels. Previous research showed that ratings of literariness are predictable from texts to a substantial extent using machine learning, suggesting that it may be possible to explain the consensus among readers on which novels are literary as a consensus on the kind of writing style that characterizes literature. Although we have not yet collected human judgments to establish the influence of writing style directly (we use a survey with judgments based on the titles of novels), we can try to analyze the behavior of machine learning models on particular text fragments as a proxy for human judgments. In order to explore aspects of the texts associated with literariness, we divide the texts of the novels in chunks of 2–3 pages and create vector space representations using topic models (Latent Dirichlet Allocation) and neural document embeddings (Distributed Bag-of-Words Paragraph Vectors). We analyze the semantic complexity of the novels using distance measures, supporting the notion that literariness can be partly explained as a deviation from the norm. Furthermore, we build predictive models and identify specific keywords and stylistic markers related to literariness. While genre plays a role, we find that the greater part of factors affecting judgments of literariness are explicable in bag-of-words terms, even in short text fragments and among novels with higher literary ratings. The code and notebook used to produce the results in this paper are available at https://github.com/andreasvc/litvecspace.
Probability density approximation (PDA) is a nonparametric method of calculating probability densities. When integrated into Bayesian estimation, it allows researchers to fit psychological processes for which analytic probability functions are unavailable, significantly expanding the scope of theories that can be quantitatively tested. PDA is, however, computationally intensive, requiring large numbers of Monte Carlo simulations in order to attain good precision. We introduce Parallel PDA (pPDA), a highly efficient implementation of this method utilizing the Armadillo C++ and CUDA C libraries to conduct millions of model simulations simultaneously in graphics processing units (GPUs). This approach provides a practical solution for rapidly approximating probability densities with high precision. In addition to demonstrating this method, we fit a piecewise linear ballistic accumulator model (Holmes, Trueblood, & Heathcote, 2016) to empirical data. Finally, we conducted simulation studies to investigate various issues associated with PDA and provide guidelines for pPDA applications to other complex cognitive models.
Most research groups studying human navigational behavior with virtual environment (VE) technology develop their own tasks and protocols. This makes it difficult to compare results between groups and to create normative data sets for any specific navigational task. Such norms, however, are prerequisites for the use of navigation assessments as diagnostic tools—for example, to support the early and differential diagnosis of atypical aging. Here we start addressing these problems by presenting and evaluating a new navigation test suite that we make freely available to other researchers (https://osf.io/mx52y/). Specifically, we designed three navigational tasks, which are adaptations of earlier published tasks used to study the effects of typical and atypical aging on navigation: a route-repetition task that can be solved using egocentric navigation strategies, and route-retracing and directional-approach tasks that both require allocentric spatial processing. Despite introducing a number of changes to the original tasks to make them look more realistic and ecologically valid, and therefore easy to explain to people unfamiliar with a VE or who have cognitive impairments, we replicated the findings from the original studies. Specifically, we found general age-related declines in navigation performance and additional specific difficulties in tasks that required allocentric processes. These findings demonstrate that our new tasks have task demands similar to those of the original tasks, and are thus suited to be used more widely.
This article introduces a package developed for R (R Core Team, 2017) for performing an integrated analysis of multiple data blocks (i.e., linked data) coming from different sources. The methods in this package combine simultaneous component analysis (SCA) with structured selection of variables. The key feature of this package is that it allows to (1) identify joint variation that is shared across all the data sources and specific variation that is associated with one or a few of the data sources and (2) flexibly estimate component matrices with predefined structures. Linked data occur in many disciplines (e.g., biomedical research, bioinformatics, chemometrics, finance, genomics, psychology, and sociology) and especially in multidisciplinary research. Hence, we expect our package to be useful in various fields.
In this paper we present a pipeline for the detection of spelling variants, i.e., different spellings that represent the same word, in non-standard texts. For example, in Middle Low German texts in and ihn (among others) are potential spellings of a single word, the personal pronoun ‘him’. Spelling variation is usually addressed by normalization, in which non-standard variants are mapped to a corresponding standard variant, e.g. the Modern German word ihn in the case of in. However, the approach to spelling variant detection presented here does not need such a reference to a standard variant and can therefore be applied to data for which a standard variant is missing. The pipeline we present first generates spelling variants for a given word using rewrite rules and surface similarity. Afterwards, the generated types are filtered. We present a new filter that works on the token level, i.e., taking the context of a word into account. Through this mechanism ambiguities on the type level can be resolved. For instance, the Middle Low German word in can not only be the personal pronoun ‘him’, but also the preposition ‘in’, and each of these has different variants. The detected spelling variants can be used in two settings for Digital Humanities research: On the one hand, they can be used to facilitate searching in non-standard texts. On the other hand, they can be used to improve the performance of natural language processing tools on the data by reducing the number of unknown words. To evaluate the utility of the pipeline in both applications, we present two evaluation settings and evaluate the pipeline on Middle Low German texts. We were able to improve the F1 score compared with previous work from \(0.39\) to \(0.52\) for the search setting and from \(0.23\) to \(0.30\) when detecting spelling variants of unknown words.
We tested the hypothesis that phonosemantic iconicity––i.e., a motivated resonance of sound and meaning––might not only be found on the level of individual words or entire texts, but also in word combinations such that the meaning of a target word is iconically expressed, or highlighted, in the phonetic properties of its immediate verbal context. To this end, we extracted single lines from German poems that all include a word designating high or low dominance, such as large or small, strong or weak, etc. Based on insights from previous studies, we expected to find more vowels with a relatively short distance between the first two formants (low formant dispersion) in the immediate context of words expressing high physical or social dominance than in the context of words expressing low dominance. Our findings support this hypothesis, suggesting that neighboring words can form iconic dyads in which the meaning of one word is sound-iconically reflected in the phonetic properties of adjace)
To qualitative researchers, social media offers a novel opportunity to harvest a massive and diverse range of content without the need for intrusive or intensive data collection procedures. However, performing a qualitative analysis across a massive social media data set is cumbersome and impractical. Instead, researchers often extract a subset of content to analyze, but a framework to facilitate this process is currently lacking. We present a four-phased framework for improving this extraction process, which blends the capacities of data science techniques to compress large data sets into smaller spaces, with the capabilities of qualitative analysis to address research questions. We demonstrate this framework by investigating the topics of Australian Twitter commentary on climate change, using quantitative (non-negative matrix inter-joint factorization; topic alignment) and qualitative (thematic analysis) techniques. Our approach is useful for researchers seeking to perform qualitative analyses of social media, or researchers wanting to supplement their quantitative work with a qualitative analysis of broader social context and meaning.
We explore here the application of modern computer hardware and software to the collection and analysis of behavioral data. We discuss the issues of ecological validity, storage and processing, data permanence, automation, validity, and algorithmic determinism. Taking the modern landscape into account, we demonstrate several varying projects we have recently undertaken as proofs of concept of the viability and utility of this approach. In particular, we describe four research projects, which involve work on child-directed speech; the application of automatic methods to clinical populations, including children with hearing loss; quality control and the assessment of validity; and the sharing of data in a public database. We conclude by pointing out how the methodology described here can be extended to a wide variety of interdisciplinary and detailed projects that are likely to lead to better science and improved outcomes for populations served by the behavioral, social, and health sciences.
THE PUBLICATION OF THE DOUGLAS FIR Group's (DFG, 2016) article reminded me of recurring paradigm debates in our field over the last 3 decades (e.g., Firth & Wagner, 1997, 2007; Zuengler & Miller, 2006). The essence of these debates revolves around the critical question of whether or not we should attribute primacy to the individual brain or the social world when exploring language learners’ learning processes. By their very nature, these paradigm debates are serious and have significant implications for research and practice, as they help to define what counts as legitimate second language acquisition (SLA) research. My colleagues often call the paradigms in these debates “bubbles,” which both envelop us and enable our intellectual pursuits. We pursue our intellectual interests through the bubble surrounding us, which often deflects reality into constructions. As a language teacher educator, I believe that the constructions in these paradigms should be put to the test by language learners and teachers in practice, so that we may recognize whether they constitute sensible solutions or responses to the challenges we face in an increasingly diverse world. For this reason, I believe that the DFG's proposal has significant implications for language teacher education, since it recommends “practical, innovative, and sustainable solutions that are responsive to the challenges of language teaching and learning in a multilingual world” (p. 20). In this commentary, I will first explain how my intellectual efforts so far have been related to the basic tenets in the DFG framework. Then I discuss how the framework, enriched by the collection of papers in this volume, presents both challenges and learning opportunities for language teachers. I conclude that transforming these challenges into learning opportunities will necessitate substantial reforms in many language teacher education programs. I welcome the publication of the DFG article on a transdisciplinary framework for SLA in a multilingual world because such a framework fits closely with the intellectual journey I myself have undertaken since my doctoral studies some years ago. Starting my academic career as a language learning strategy researcher, I have been interested not so much in strategies in and of themselves but in understanding why language learners adopt particular strategies in a given context (e.g., Gao, 2010). In that sense, my research on language learning strategies has always been transdisciplinary, drawing on sociocultural perspectives on language learning and debates about agency and structure in the social sciences (Layder, 1990) to inform my interpretation of learners’ strategic language learning narratives. In my research I portray language learners’ strategy use not only as a cognitive choice, but also as a choice facilitated or constrained by contextual conditions (e.g., Gao, 2007, 2010). To some extent, then, my multifaceted approach to language learners’ strategy use echoes the theoretical considerations captured by the DFG framework while applying them to a particular area of inquiry. At the same time, it is fair to say that, although the publication of my doctoral thesis as a research monograph was positively reviewed by some researchers in the field, overall, its ideas have not received much attention in the academic community. In part, this may be due to the fact that language learning strategy researchers face an almost insurmountable challenge posed by those advocating the replacement of the construct with self-regulated learning capacity (e.g., Rose et al., 2018; Tseng, Dörnyei, & Schmitt, 2006). But much larger issues are at play: Socializing with SLA researchers at major applied linguistics conferences frequently leaves me with the impression that my approach, at best, is alternative to more mainstream and dominant approaches, perhaps even marginal. Fortunately, after I became a language teacher educator that discomfort has gradually subsided—most likely because my new role challenged me to commit myself to conducting research that spoke more directly to the concerns of language teachers. Immersing myself in research with explicit pedagogical implications, I realized that language learning and teaching take place in a complex milieu, mediated by a variety of conditions and relationships within and beyond the confines of classroom walls. It is well known that many of these are beyond the teachers’ control. And yet, having been socialized into a professional culture that has long assigned responsibility for learner success to them, they are eager to find ‘best practices’ that would assure that success. With regard to motivation, for example, language teachers often look to teacher educators for ‘magic’ ways through which they might ‘trigger’ a redoubling of efforts on the part of their learners. However, even though motivation is indeed a key factor in explaining variations in learners’ achievements (Boo, Dörnyei, & Ryan, 2015), motivation is an extraordinarily multifaceted phenomenon, arising from the interaction between individual language learners’ beliefs and attitudes, input and influences from social agents (e.g., peers, parents, teachers) within the different social groups (or communities) to which they belong, as well as ideological discourses about learning at the macro-contextual level (Gao, 2010; Harvey, 2017; Lanvers, 2018). As I stated at the outset, discussions of paradigm shifts are hardly new in the language studies field. Even so, I welcome the DFG framework with its 10 research themes as a key intellectual resource underlying transdisciplinary, innovative research that can generate critical insights for language learners and teachers to improve their well-being. That positive assessment may, in part, stem from the fact that the DFG framework directly connects my work to some of the central debates in our field, rather than relegating it to the periphery. Far more important, however, is my belief that the DFG framework can serve as a pedagogical resource for language teachers thatenables them to position their teaching within the big picture of learning and teaching and, thereby, encourage them to re-commit themselves to supporting their own language learners by any means whatsoever. It also provides a guide for language teacher educators to help language teachers maneuver strategically in pursuit of professional development. In other words, the framework not only presents a research resource map that underpins holistic inquiries into “language learning and teaching” by “taking into account forces beyond individual learners” (DFG, 2016, p. 20); it also maps out the complexity of the learning and teaching that language teacher educators need to impart to language teachers for their professional development. The collection of studies in this issue can be approached from different perspectives, but as a language teacher educator I believe that they highlight at least four challenges for teacher educators in the preparation of language teachers: (a) the deficit discourses about language learners associated with the monolingualism bias in SLA research (Ortega, 2019, this issue), (b) integrating a broadened theorization of cognition in teaching (Ellis, 2019, this issue), (c) taking a usage-based approach to linguistics to inform pedagogical decisions (Hall, 2019, this issue; LaScotte & Tarone, 2019, this issue), and (d) becoming fully empowered agents in the pedagogical process (Larsen–Freeman, 2019, this issue). The first three challenges raise critical questions about why language teachers teach, how language learners learn (and hence how language can be taught), and what is to be taught. Properly addressed, language teacher education programs will present rich, meaningful learning opportunities that are essential to language teachers’ professional development as they learn the task of teaching in a multilingual world. The last challenge goes to the heart of language teachers’ professional development as the DFG framework envisions it. One of the most significant challenges in my career as a language teacher educator has been to help language teachers recognize the deleterious consequences of the reigning monolingual prejudice in SLA research, which has led to the enshrining of native speaker norms as the privileged target of all language learning and teaching. Perhaps such a goal is achievable by a select few learners in highly unusual circumstances and, most likely, even then at considerable costs (e.g., financial and social). But its dominance (for an example in the Chinese context, see Wang, 2015) based on a fundamental bias in favor of a small group of privileged language learners is extraordinarily difficult to address—much less resist—for researchers, teacher educators, teachers, and learners alike. Research conducted in Western,educated, industrialized, rich, and democratic contexts (“WEIRD” research) prides itself on producing complex and elaborate justifications and procedures that are intended to inform language teachers’ pedagogical decisions. But teachers in other contexts may find their arguments as well as their educational recommendations both unpersuasive and irrelevant, not least because of the glaring gap between the biases of such academic research and their own pedagogical reality. Even more important, by reducing the complex reality of multilingual competence to dichotomies like native and nonnative, monolingual and bilingual, they inflict additional, deeply disturbing insecurities beyond those that are already part of the complex endeavor of language learning and language use—and do so for both the multilingual learners as well as their teachers (cf. Ortega, 2019, this issue). Over the years I have found that it is ineffective to address the impact of deficit discourses on language teachers simply by changing the ways in which we describe language learners and learning, for example by changing “language barrier” to “language difference.” Without a fundamental shift in how teachers can understand the verynature of their professional work and, accordingly, the nature of their commitments and responsibilities, well-intentioned replacive words with positive meanings all too soon assume the same negative connotations that have pervaded the earlier deficit discourses. As suggested by Ortega, only when language teacher education programs incorporate social justice and equity into their core missions can we begin to tackle this deep-seated bias. That means, language teacher educators need more research that pertains to a myriad of aspects of language use and language learning as these describe the vast majority of language learners who do not belong to the over-studied multilingual elites. One obvious starting point for such research might be the multilingual teachers themselves who, around the globe, bring particular strengths to their work and—as we hear over and over again from their students—are role models that inspire and motivate, often for a lifetime. For years, language teachers have been urged to provide “authentic language” through “natural learning” (e.g., Widdowson, 1998, p. 705; original italic). However, language teachers “need to make language and language learning a reality for learners” (Widdowson, 1998, p. 715), a demand that draws attention to theorization about language cognition in context. By focusing on language learning within micro and macro contexts, the DFG framework presents a theorization of cognition as socially distributed and as taking place within a human body that connects cognition and the world. Individual learners’ language cognition is dependent on context and is enacted through learner action upon the context (e.g., Atkinson et al., 2018; Zuengler & Miller, 2006). This recognition underpins the growth of a variety of theories of language learning, including sociocultural perspectives (e.g., Lantolf, 2009), complex dynamic system theory (e.g., Larsen–Freeman, 1997), and usage-based approaches to language development (e.g., Ellis & Larsen–Freeman, 2006). With regard to language pedagogy development, these perspectives on language learning imply the necessity for language teachers to draw on what language learners are exposed to outside the classroom in order to develop meaningful pedagogical activities in the classroom, which will prepare them well for the task of using the language in real life. In other words, language teachers need to consider “frequent and rich participation in the second-language life worlds into which the learner ‘bricolages’ his or her way” (Ellis, 2019, this issue, p. 52) as both the starting and end points for language teaching activities within the classroom toward language learning and use. Considering the fact that many language learners (particularly those in low-resource situations) have few opportunities to use the language outside the classroom, it remains a critical task for language teachers to develop, create, and sustain learners’ meaningful access to these “second-language life worlds,” not least because doing so may help reduce inequity among language learners in the learning process. However, most language teacher education programs that I am aware of have yet to develop language teachers’ capacity to promote language use beyond the classroom in low-resource teaching contexts. The socially distributed nature of cognition, as outlined by Ellis, draws attention to the significance of the gaps that different language learners may experience in their exposure to the target language and their ultimate language learning achievement. This requires language teacher educators and language teachers to work out ways to address these gaps and minimize the impact of social inequity on language learners’ learning progress. The usage-based understanding of language, as one fundamental pillar of the DFG framework, presents another significant challenge for language teachers and language teacher educators. For years, we have engaged in debating form versus meaning in language teaching. But as Hall (2019, this issue) argues, a usage-based understanding of language necessitates fundamental shifts in conceptualizing a number of key constructs in language teaching, such as replacing “language competence” with “language repertoire,” and “language expertise” as well as “grammar” with “semiotic resources” and “semioticregisters.” Drawing on interactional linguistics, Hall (2019, this issue) presents grammatical practices as “shaped by actional turns…within interactional sequences” (p. 84), and conversation analysis as seeking to uncover “the publically witnessable universal infrastructure, as exhibited in the methods used by members to achieve social order” (p. 86). In light of such an understanding of language, many language teacher education programs will need to examine critically how they theorize language so that the programs can prepare language teachers with “theories of language that foreground meaning rather than structure” (Lantolf, 2009, p. 274). Among the consequences of such an approach will be that language teachers should encourage learners to deploy lexical and grammatical resources from different registers to establish their own voices or enact those of others, as revealed in LaScotte and Tarone's (2019, this issue) study on language learners’ differential development of complexity, accuracy, and fluency in different ‘voice’ environments. In sum, the articles in this collection present a coherent transdisciplinary framework, which entails significant challenges for language teachers and language teacher educators in the pursuit of a more equitable world through language education. To overcome these challenges, we must recognize language teachers as reflexive and reflective agents, who are able to critically evaluate knowledge resources and uncertain conditions before making pedagogical decisions that advance their professional ideals. That means, we need to understand how teacher agency underlies professional development. Larsen–Freeman (2019, this issue) enriches the Douglas Fir Group framework with a temporal element with regard to learner agency. That enables her to capture the dynamism of the framework as compared with the earlier, more flat and two-dimensional portrayal. Language teachers’ professional development, too, can be seen as a dynamic process emerging from interaction between teacher agency and contextual structures as outlined in the framework. While the DFG framework can be used as a resource to help language teachers appreciate the complexity and dynamism of language learning and teaching, it is also a roadmap for them to see how their professional learning might unfold in the specific contexts where they learn and teach. Depending on where we locate agency in the framework, agency can be theorized as capacity or phenomenon/doing (Priestley et al., 2012). If we place agency with “individuals engaging with others” in the framework, we may consider it as “the socio-culturally mediated capacity to act purposefully and reflectively on [one's] world” (Rogers & Wetzel, 2013, p. 63; see also Ahearn, 2001). Such a consideration acknowledges social contributions to the development and exercise of agency, but it may not be well suited to helping us to “identify agentic actions,” as agency and structure are mutually constitutive in this conceptualization (Hitlin & Elder, 2007, p. 173). For this reason, it is, perhaps, more appropriate to regard language teachers’ agency as phenomenon/doing, or something “achieved and not as merely … a capacity or possession of the individual” (Priestley et al., 2012, p. 197), a position that confirms the significance of adding Larsen–Freeman's (2019, this issue) temporal element to the DFG framework. Inevitably, its integration into language teacher education programs compels language teachers to critically examine and change their beliefs about language, learning, and teaching. The traditional emphasis on language structure in language teacher education programs does not help language teachers to prepare learners for using language structures as semiotic resources to create meanings and assert themselves. Rather, it creates an irresistibly tempting—though false—sense of security and authority for language teachers. Without the aid of deterministic language rules and orderly language development paths as these have been suggested by many SLA studies, language teachers will be called to “confront, resist and work out pedagogical conflicts” as well as “overcome unequal power relations” in professional practice, in which “teacher agency plays a key role” (Tao & Gao, 2017, p. 347). That means that we must recognize their “rights” to control and their “responsibility” to sustain their professional development for the task of language teaching, positions that are reflected in the DFG framework (see also, Johnson, 2006, p. 250). Since language teachers learn and teach in a broader milieu of preset conditions and prior experiences in particular contexts, their agentic responses may be manifest in the forms of compliance, resistance, and negotiation over the course of time. Language teacher educators and researchers need to critically examine and endeavor to change structural elements (e.g., conceptions of language learning and teaching endorsed by the powerful) that might have constrained language teachers’ learning and the application of new pedagogical ideas in a given teaching context. In other words, language teacher educators not only need to provide knowledge and experience to enable transformation among language teachers; they should also explore ways to create and sustain the contextual conditions that are conducive to changes in their learning and professional practice. Publication of the DFG (2016) article may be regarded as a in the development of what counts as SLA research. The framework can be seen as a point for a group of SLA researchers who on related to language teaching In my it also language teacher educators like me to how the framework can be used to inform the development of language teacher education programs by why we teach what we teach, and how we teach Language teacher education programs should on language teachers who are to social justice and equity in teaching, who are critically aware of the significant impact that contextual conditions may have on language learners’ learning, and who can help language learners develop semiotic resources to assert themselves in contexts. central tenets of dynamic complex system theory with a rich understanding of teacher agency may prepare language teacher educators and language teachers for a understanding of their own professional The DFG framework can also as a map to guide language teachers’ efforts to prepare themselves for the task of learning to create a more equitable world through language teaching. of us a difficult process of in which agency will a critical role in transforming our understanding and our pursuit of professional For the of a more equitable we must and
Previous research findings supporting the advantages of the go/no-go choice over the yes/no choice in lexical decision task (LDT) have suggested that the go/no-go choice might require less cognitive resources in the non-decisional processes. This study aims to test such an idea using the event-related potential method. In this study, the tasks (yes/no LDT and go/no-go LDT) and word frequency (high and low) were manipulated, and the difference between the go/no-go choice and yes/no choice were examined with BP, pN, pN1, P200, N400, and P3 components that were assumed to be closely related with the various parameters in the diffusion model. The results showed that BP, pN and pN1 amplitudes reflecting the preparation stage were not differently affected by word frequency and the task type. However, ERPs after stimulus onset showed differences. The P200 amplitudes were smaller in the go/no-go task than in the yes/no task only for low-frequency words. N400 and P3 amplitudes were only affect)
This study proposes and evaluates a general diagnostic classification model (DCM) for rating scales. We applied the proposed model to a dataset to compare its performance with traditional DCMs for polytomous items. We also conducted a simulation study based on the applied study condition in order to evaluate the parameter recovery of the proposed model. The findings suggest that the proposed model shows promise for (1) accommodating much smaller sample sizes by reducing a large number of parameters for estimation; (2) obtaining item category response probabilities and individual scores very similar to those from a traditional saturated model; and (3) providing general item information that is not available in traditional DCMs for polytomous items.
With the development of online data collection and instruments such as Amazon’s Mechanical Turk (MTurk), the appearance of malicious software that generates responses to surveys in order to earn money represents a major issue, for both economic and scientific reasons. Indeed, even if paying one respondent to complete one questionnaire represents a very small cost, the multiplication of botnets providing invalid response sets may ultimately reduce study validity while increasing research costs. Several techniques have been proposed thus far to detect problematic human response sets, but little research has been undertaken to test the extent to which they actually detect nonhuman response sets. Thus, we proposed to conduct an empirical comparison of these indices. Assuming that most botnet programs are based on random uniform distributions of responses, we present and compare seven indices in this study to detect nonhuman response sets. A sample of 1,967 human respondents was mixed with different percentages (i.e., from 5% to 50%) of simulated random response sets. Three of the seven indices (i.e., response coherence, Mahalanobis distance, and person–total correlation) appear to be the best estimators for detecting nonhuman response sets. Given that two of those indices—Mahalanobis distance and person–total correlation—are calculated easily, every researcher working with online questionnaires could use them to screen for the presence of such invalid data.
The human task-evoked pupillary response provides a sensitive physiological index of the intensity and online resource demands of numerous cognitive processes (e.g., memory retrieval, problem solving, or target detection). Cognitive pupillometry is a well-established technique that relies upon precise measurement of these subtle response functions. Baseline variability of pupil diameter is a complex artifact that typically necessitates mathematical correction. A methodological paradox within pupillometry is that linear and nonlinear forms of baseline scaling both remain accepted baseline correction techniques, despite yielding highly disparate results. The task-evoked pupillary response (TEPR) could potentially scale nonlinearly, similar to autonomic functions such as heart rate, in which the amplitude of an evoked response diminishes as the baseline rises. Alternatively, the TEPR could scale similarly to the cortical hemodynamic response, as a linear function that is independent of its baseline. However, the TEPR cannot scale both linearly and nonlinearly. Our aim was to adjudicate between linear and nonlinear scaling of human TEPR. We manipulated baseline pupil size by modulating the illuminance in the testing room as participants heard abrupt pure-tone transitions (Exp. 1) or visually monitored word lists (Exp. 2). Phasic pupillary responses scaled according to a linear function across all lighting (dark, mid, bright) and task (tones, words) conditions, demonstrating that the TEPR is independent of its baseline amplitude. We discuss methodological implications and identify a need to reevaluate past pupillometry studies.
Even though human behavior is largely driven by real-time feedback from others, this social complexity is underrepresented in psychological theory, largely because it is so difficult to isolate. In this work, we performed a quasi-experimental analysis of hundreds of millions of chat room messages between young people. This allowed us to reconstruct how—and on what timeline—the valence of one message affects the valence of subsequent messages by others. For the highly emotionally valenced chat messages that we focused on, we found that these messages elicited a general increase of 0.1 to 0.4 messages per minute. This influence started 2 s after the original message and continued out to 60 s. Expanding our focus to include feedback loops—the way a speaker’s chat comes back to affect him or her—we found that the stimulating effects of these same chat events started rippling back from others 8 s after the original message, to cause an increase in the speaker’s chat that persisted for up to 8 min. This feedback accounted for at least 1% of the bulk of chat. Additionally, a message’s valence affects its dynamics, with negative events feeding back more slowly and continuing to affect the speaker longer. By reconstructing the second-by-second dynamics of many psychosocial processes in aggregate, we captured the timescales at which they collectively ripple through a social system to drive system-level outcomes.
Discourse markers are words and expressions (such as: firstly, then, for example, because, as a result, likewise, in comparison, in contrast) that explicitly state the relational structure of the information in the text, i.e. signalling a sequential relationship between the current message and the previous discourse. Using these markers improves the cohesion and coherence of texts, facilitating reading comprehension. Although often included in tools that support the rhetoric structuring of texts, discourse markers have hardly been explored in writing support tools for learners of a second language. However, learners of a second language, including those at advanced levels, have trouble producing these lexical items, frequently replacing them with items from their native language or with literal translations of items in their own language, which often do not result in proper lexical items in the second language. In addition, students learn a single marker per function and use it repeatedly, producing monotonous texts. With the aim of contributing to reducing these difficulties, this paper presents a lexicon that will be used to support the task of automatically detecting and correcting discourse marker errors. Several heuristics have been evaluated to generate different types of errors. Automatic translation methods were used to semi-automatically compile the lexicon used in these heuristics. Similarity measures were also combined with these heuristics to correct discourse marker errors. The evaluated methods proved to be suitable for the task of identifying some types of discourse marker errors and can potentially identify many others, as long as new lexical inputs are incorporated into them.
The notion of spreading activation is a central theme in the cognitive sciences; however, the tools for implementing spreading activation computationally are not as readily available. This article introduces the spreadr R package, which can implement spreading activation within a specified network structure. The algorithmic method implemented in the spreadr subroutines follows the approach described in Vitevitch, Ercal, and Adagarla (Frontiers in Psychology, 2, 369, 2011), who viewed activation as a fixed cognitive resource that could “spread” among connected nodes in a network. Three sets of simulations were conducted using the package. The first set of simulations successfully reproduced the results reported in Vitevitch et al. (Frontiers in Psychology, 2, 369, 2011), who showed that a simple mechanism of spreading activation could account for the clustering coefficient effect in spoken word recognition. The second set of simulations showed that the same mechanism could be extended to account for higher false alarm rates for low clustering coefficient words in a false memory task. The final set of simulations demonstrated how spreading activation could be applied to a semantic network to account for semantic priming effects. It is hoped that this package will encourage cognitive and language scientists to explicitly consider how the structures of cognitive systems such as the mental lexicon and semantic memory interact with the process of spreading activation.
Severe developmental deficits in face recognition ability (developmental prosopagnosia, or DP) have been vigorously studied over the past decade, yet many questions remain unanswered about their origins, nature, and social consequences. A rate-limiting factor in answering such questions is the challenge of recruiting rare DP participants. Although self-reported experiences have long played a role in efforts to identify DPs, much remains unknown about how such self-reports can or should contribute to screening or diagnosis. Here, in a large, population-based web sample, we investigated the effectiveness of self-report, used on its own, as a screen to identify individuals who will ultimately fail, at a conventional cutoff, the two types of objective tests that are most commonly used to confirm DP diagnoses: the Cambridge Face Memory Test (CFMT) and the famous faces memory test (FFMT). We used a highly reliable questionnaire (alpha = .91), the Cambridge Face Memory Questionnaire (CFMQ), and revealed strong validity via high correlations of .44 with the CFMT and .52 with the FFMT. However, cutoff analyses revealed that no CFMQ score yielded a clinical-grade combination of sensitivity and positive predictive value in enough individuals to support using it alone as a DP diagnostic or screening tool. This result was replicated in an analysis of data from the widely used PI20 questionnaire, a 20-question self-assessment of facial recognition similar in form to the CFMQ. We therefore recommend that screens for DP should, wherever possible, include objective as well as subjective assessment tools.
The article deals with the semantics of the term “populism” and its re-lexicalization via changing the meaning from neutral to negative. The author gives examples of the third-wave American populist leaders’ speeches, namely those made by Ugo Chaves, Nicolas Maduro, Evo Morales and Daniel Ortega, and reveals linguistic means of populism, as well as discursive strategies: proliferation (lexical loading) of key concepts, such as “pueblo”, “nación”, “patria” (“people”, “nation”, “motherland”), constructing discursive “I” in a pattern “I am my people”, the use of discursive strategies of polarization that helps political leaders delegitimize the opponent and construct the para-reality to legalize and symbolize self-power. The populist discourse is marked semantically by key words pueblo, nación, patria, special contextual coloring is designed with set expressions like la voz del pueblo (voice of the people), by forming occasional derivatives that break language norms, by actualizing oppositions “friend - enemy”, “acquaintance – stranger”, “people – outsider”, by applying to slogans and precedent phenomena. The participation of precedent phenomena in discursive strategies of populism is highlighted in the choice of precedent names (i.e. Simón Bolívar, Augusto César Sandino, Rubén Darío), it is directed at affiliating modern political leaders with prominent leader of the past, thus pointing to national values and providing legitemisy of the politicains in power.
INTRODUCTION: The aim of the present study was to investigate how emotions influence pain, measured by one subjective self-rated measure, the numeric rating scale (NRS), and one objective physiological measure, the number of skin conductance responses (NSCR). METHOD: Eighteen volunteers were exposed to conditions with pictorial emotional stimuli (neutral, positive, negative), authentic ICU-sound (noise, no-noise) and electrical stimulation (pain, no-pain) individually titrated to induce moderate pain. When using all combinations of picture inducing emotions, sound, and pain, each of these conditions (12 conditions lasting for 60 seconds each) were followed by pain ratings. Ratings of arousal (low to high) and valence (pleasant to unpleasant) were used as indicators of affective state for each condition. Mean NSCR was also measured throughout the experiment for each condition. RESULTS: Even though NRS and NSCR increased during painful stimuli, they did not correlate during the trial. However, NSCR was positively correlated with the strength of the electrical stimulation, r = 0.48, P = 0.046, whereas NRS showed positive correlations with the anxiety level, assessed by affective ratings (arousal, r = 0.61, P < 0.001, and valence, r = 0.37, P < 0.001). CONCLUSIONS: The NRS was strongly influenced by affective state, with higher pain ratings during more anxiety-like states, whereas NSCR correlated to the strength of electrical pain stimulation. That reported pain is moderated by anxiety, puts forward a discussion whether reduction of the anxiety level should be considered during analgesia treatment.
This position paper is based on a keynote presentation at the COLING 2016 Workshop on Language Technology for Digital Humanities in Osaka, Japan. It departs from observations about working practices in Humanities disciplines following a hermeneutic tradition of text interpretation versus the method-oriented research strategies in Computational Linguistics (CL). The respective praxeological traditions are quite different. Yet more and more researchers are willing to open up towards truly transdisciplinary collaborations, trying to exploit advanced methods from CL within research that ultimately addresses questions from the traditional Humanities disciplines and the Social Sciences. The article identifies two central workflow-related issues for this type of collaborative project in the Digital Humanities (DH) and Computational Social Science: (1) a scheduling dilemma, which affects the point in the course of the project when specifications of the core analysis task are fixed (as early as possible from the computational perspective, but as late as possible from the Humanities perspective) and (2) the subjectivity problem, which concerns the degree of intersubjective stability of the target categories of analysis. CL methodology demands high inter-annotator agreement and theory-independent categories, while the categories in hermeneutic reasoning are often tied to a particular interpretive approach (viz. a theory of literary interpretation) and may bear a non-trivial relation to a reader’s pre-understanding. Building a comprehensive methodological framework that helps overcome these issues requires considerable time and patience. The established computational methodology has to be gradually opened up to more hermeneutically oriented research questions; resources and tools for the relevant categories of analysis have to be constructed. This article does not call into question that well-targeted efforts along this path are worthwhile. Yet, it makes the following additional programmatic point regarding directions for future research: It might be fruitful to explore—in parallel—the potential lying in DH-specific variants of the concept of rapid prototyping from Software Engineering. To get an idea of how computational analysis of some aspect of text might contribute to a hermeneutic research question, a prototypical analysis model is constructed, e.g., from related data collections and analysis categories, using transfer techniques. While the initial quality of analysis may be limited, the idea of rapid probing allows scholars to explore how the analysis fits in an actual workflow on the target text data and it can thus provide early feedback for the process of refining the modeling. If the rapid probing method can indeed be incorporated in a hermeneutic framework to the satisfaction of well-disposed Humanities scholars, a swifter exploration of alternative paths of analysis would become possible. This may generate considerable additional momentum for transdisciplinary integration. It is as yet too early to point to truly Humanities-oriented examples of the proposed rapid probing technique. To nevertheless make the programmatic idea more concrete, the article uses two experimental scenarios to argue how rapid probing might help addressing the scheduling dilemma and the subjectivity problem respectively. The first scenario illustrates the transfer of complex analysis pipelines across corpora; the second one addresses rapid annotation experiments targeting character mentions in literary text.
Drawing on Gilles Deleuze and Félix Guattari’s analysis of language and literature, formulated in Kafka: pour une littérature mineure (1975), Mille plateaux (1980) and elsewhere, this article pursues the idea that certain forms of language, such as national languages or major literary discourses, might be conceptualized as states. Each operating as a locus of power, these forms of language codify regulatory rules (grammatical, lexical or stylistic, for example) that serve to stake out the boundaries of their territories; adherence to these rules and norms subsequently identifies belonging or non-belonging to a given linguistic or literary community. Concomitant with the notion of the linguistic state is the notion of linguistic statelessness, which might describe a general sense of alienation within language, such as that produced by the defamiliarizing effects of poetic discourse, or a more localized sense, prompted by the socio-political situation of certain marginalized, regional languages. Taking Deleuze and Guattari’s notion of ‘déterritorialisation’ as a point of departure, this article explores how certain forms of literature dismantle or disrupt dominant linguistic codes, pointing to a decentralized position outside of the established state. Here, literature involves a movement outside of a given linguistic territory, prompting a kind of statelessness within language. Having elaborated these notions of ‘state’ and ‘statelessness’, ‘territories’ and ‘deterritorialisation’ in greater detail, the article traces their configuration in the work of the contemporary French poet, Olivier Cadiot. It considers Cadiot’s first collection of poetry, L’art poétic’ (1988), and then his collaboration with the musician Rodolphe Burger on the album Welche: On n’est pas indiens c’est dommage (2000). In both instances, Cadiot uses ready-made language, employing cut-ups and sampling techniques that rework dominant discourses, deterritorializing them and making them ‘minor’.
This paper presents a combination of R packages—user contributed toolkits written in a common core programming language—to facilitate the humanistic investigation of digitised, text-based corpora. Our survey of text analysis packages includes those of our own creation (cleanNLP and fasttextM) as well as packages built by other research groups (stringi, readtext, hyphenatr, quanteda, and hunspell). By operating on generic object types, these packages unite research innovations in corpus linguistics, natural language processing, machine learning, statistics, and digital humanities. We begin by extrapolating on the theoretical benefits of R as an elaborate gluing language for bringing together several areas of expertise and compare it to linguistic concordancers and other tool-based approaches to text analysis in the digital humanities. We then showcase the practical benefits of an ecosystem by illustrating how R packages have been integrated into a digital humanities project. Throughout, the focus is on moving beyond the bag-of-words, lexical frequency model by incorporating linguistically-driven analyses in research.
espanolEn 1969, Philippe Jaccottet tradujo para la revista Aquila el poema Dunja de Giuseppe Ungaretti. El estudio tiene como objetivo destacar la singularidad de esta traduccion, arrojar luz sobre la genetica de su escritura: es decir, penetrar el lado invisible del proceso de traduccion en el que participan el autor y el traductor. A partir de la aclaracion terminologica de la nocion de transparencia, se aplica para dilucidar las estrategias del traductor, destacando la dialectica de la hermetica y de la hermeneutica. La yuxtaposicion de las versiones del poema, elaboradas por Jaccottet durante el trabajo en curso, proporciona informacion sobre los pasos de la traduccion transparente, comenzando por el trabajo sobre el lexico que se refiere a la busqueda de la equivalencia. Revela, ademas, las indecisiones y las dudas de un traductor advertido, consciente de la imposibilidad de comprender el enunciado poetico, descuidando sus aspectos emocional, imaginal y prosodico. Para llegar a la conclusion de que se trata de una traduccion “relevante”, resultado de un trabajo meticuloso sobre las potencialidades de la lengua meta, que no duda en transgredir sus normas gramaticales para preservar la extraneza del texto fuente. EnglishIn 1969 Philippe Jaccottet translated for the magazine Aquila the poem Dunja by Giuseppe Ungaretti. The study aims to highlight the singularity of this translation, shedding light on the genetics of its writing: that is to say to penetrate the invisible side of translating in which both the author and the translator participate. Starting from the terminological clarification of the notion of transparency, it applies itself to elucidate the translator’s strategies by highlighting the dialectic of hermetics and hermeneutics. The juxtaposition of the poem’s versions, elaborated by Jaccottet during the work in progress, provides information on the stages of transparent translation starting with the lexical work which concerns the search for equivalence. It reveals furthermore the hesitations and doubts of a wise translator, aware of the impossibility of understanding the poetic statement by neglecting its emotional, imaginal and prosodic aspects. In order to reach the conclusion that this is indeed a “relevant” translation, resulting from a meticulous work on the potentialities of the target language, which does not hesitate to transgress its grammatical norms with the purpose of preserving the foreignness of the source text. francaisEn 1969, Philippe Jaccottet traduit pour la revue Aquila le poeme Dunja de Giuseppe Ungaretti. L’etude tente de relever la singularite de cette traduction, en jetant de la lumiere sur la genetique de sa redaction: autrement dit, de penetrer le cote invisible du processus traductif auquel collaborent l’auteur et le traducteur. A partir de la clarification terminologique de la notion de transparence, elle s’applique a elucider les strategies du traducteur, en mettant en relief la dialectique de l’hermetique et de l’hermeneutique. La juxtaposition des versions jaccottiennes du poeme, elaborees au cours du « work in progress », renseigne sur les etapes de la traduction transparente, a commencer par le travail sur le lexique qui concerne la recherche d’equivalence. Elle revele en outre les hesitations et les doutes d’un traducteur averti, conscient de l’impossibilite de comprendre l’enonce poetique, en negligeant ses aspects affectif, imaginal et prosodique. Pour aboutir a la conclusion qu’il s’agit bien d’une traduction « relevante », resultat d’un travail minutieux sur les potentialites de la langue cible, qui n’hesite pas a transgresser les normes grammaticales de celle-ci afin de preserver l’etrangete du texte source.
Abstract A number of longitudinal studies of L2 production have reported frequency effects wherein learners' produce more frequent words as a function of time. The current study investigated the spoken output of English L2 learners over a four-month period of time using both native and non-native English speaker frequency norms for both word types and word tokens. The study also controlled for individual differences such as first language distance, English proficiency, gender, and age. Results demonstrated that lower level L2 learners produced more infrequent tokens at the beginning of the study and that high intermediate learners, when compared to advanced learners, produced more infrequent tokens at the beginning of the study and more frequent tokens toward the end of the study. Main effects were also reported for proficiency level, age, and language distance. These results provide further evidence that L2 production may not follow expected frequency trends (i.e., that more infrequent tokens are produced as a function of time).
Empirical methods in geoparsing have thus far lacked a standard evaluation framework describing the task, metrics and data used to compare state-of-the-art systems. Evaluation is further made inconsistent, even unrepresentative of real world usage by the lack of distinction between the different types of toponyms, which necessitates new guidelines, a consolidation of metrics and a detailed toponym taxonomy with implications for Named Entity Recognition (NER) and beyond. To address these deficiencies, our manuscript introduces a new framework in three parts. (Part 1) Task Definition: clarified via corpus linguistic analysis proposing a fine-grained Pragmatic Taxonomy of Toponyms. (Part 2) Metrics: discussed and reviewed for a rigorous evaluation including recommendations for NER/Geoparsing practitioners. (Part 3) Evaluation data: shared via a new dataset called GeoWebNews to provide test/train examples and enable immediate use of our contributions. In addition to fine-grained Geotagging and Toponym Resolution (Geocoding), this dataset is also suitable for prototyping and evaluating machine learning NLP models.
Is phonetic information encoded by distributional biases in the lexicon? Are phonotactic constraints robust enough to help a learner infer the phonic pattern of a language? Our work in progress attempts to shed light on these questions via: (i) statistical description of the distributional biases in phone sequences in a lexical database and (ii) connectionist simulation. The simulation focuses on V-to-V relations in V(C)’C(C)V phone strings since both harmony and contour constraints (the tendency for the vowels to share or avoid repetition of phonic properties, respectively) have been found in the distributional study (Albano, 2002).
This paper presents experiments in part-of-speech tagging of low-resource languages. It addresses the case when no labeled data in the targeted language and no parallel corpus are available. We only rely on the proximity of the targeted language to a better-resourced language. We conduct experiments on three French regional languages. We try to exploit this proximity with two main strategies: delexicalization and transposition. The general idea is to learn a model on the (better-resourced) source language, which will then be applied to the (regional) target language. Delexicalization is used to deal with the difference in vocabulary, by creating abstract representations of the data. Transposition consists in modifying the target corpus to be able to use the source models. We compare several methods and propose different strategies to combine them and improve the state-of-the-art of part-of-speech tagging in this difficult scenario.
The transfer of research data management from one institution to another infrastructural partner is all but trivial, but can be required, for instance, when an institution faces reorganization or closure. In a case study, we describe the migration of all research data, identify the challenges we encountered, and discuss how we addressed them. It shows that the moving of research data management to another institution is a feasible, but potentially costly enterprise. Being able to demonstrate the feasibility of research data migration supports the stance of data archives that users can expect high levels of trust and reliability when it comes to data safety and sustainability.
In this paper we explain the difference between two aspects of semantic relatedness: taxonomic and thematic relations. We notice the lack of evaluation tools for measuring thematic relatedness, identify two datasets that can be recommended as thematic benchmarks, and verify them experimentally. In further experiments, we use these datasets to perform a comprehensive analysis of the performance of an extensive sample of computational models of semantic relatedness, classified according to the sources of information they exploit. We report models that are best at each of the two dimensions of semantic relatedness and those that achieve a good balance between the two.
By building a part-of-speech (POS) tagger for Middle High German, we investigate strategies for dealing with a low resource, diverse and non-standard language in the domain of natural language processing. We highlight various aspects such as the data quantity needed for training and the influence of data quality on tagger performance. Since the lack of annotated resources poses a problem for training a tagger, we exemplify how existing resources can be adapted fruitfully to serve as additional training data. The resulting POS model achieves a tagging accuracy of about 91% on a diverse test set representing the different genres, time periods and varieties of MHG. In order to verify its general applicability, we evaluate the performance on different genres, authors and varieties of MHG, separately. We explore self-learning techniques which yield the advantage that unannotated data can be utilized to improve tagging performance on specific subcorpora.
There is increasing evidence that hand gestures and speech synchronize their activity on multiple dimensions and timescales. For example, gesture’s kinematic peaks (e.g., maximum speed) are coupled with prosodic markers in speech. Such coupling operates on very short timescales at the level of syllables (200 ms), and therefore requires high-resolution measurement of gesture kinematics and speech acoustics. High-resolution speech analysis is common for gesture studies, given that field’s classic ties with (psycho)linguistics. However, the field has lagged behind in the objective study of gesture kinematics (e.g., as compared to research on instrumental action). Often kinematic peaks in gesture are measured by eye, where a “moment of maximum effort” is determined by several raters. In the present article, we provide a tutorial on more efficient methods to quantify the temporal properties of gesture kinematics, in which we focus on common challenges and possible solutions that come with the complexities of studying multimodal language. We further introduce and compare, using an actual gesture dataset (392 gesture events), the performance of two video-based motion-tracking methods (deep learning vs. pixel change) against a high-performance wired motion-tracking system (Polhemus Liberty). We show that the videography methods perform well in the temporal estimation of kinematic peaks, and thus provide a cheap alternative to expensive motion-tracking systems. We hope that the present article incites gesture researchers to embark on the widespread objective study of gesture kinematics and their relation to speech.
<strong>🌙 This is an alpha pre-release of spaCy v2.1.0 and available on pip as <code>spacy-nightly</code>. It's not intended for production use. See here for the updated nightly docs.</strong> <pre><code class="lang-bash">pip install -U spacy-nightly </code></pre> If you want to test the new version, we recommend using a new virtual environment. Also make sure to download the new models – see below for details and benchmarks. ⚠️ Due to difficulties linking our new <code>blis</code> for faster platform-independent matrix multiplication, this nightly release currently doesn't work on Python 2.7 on Windows. We expect this to be corrected in the future. ✨ New features and improvements Tagger, Parser, NER and Text Categorizer <strong>NEW:</strong> Experimental ULMFit/BERT/Elmo-like pretraining (see #2931) via the new <code>spacy pretrain</code> command. This pre-trains the CNN using BERT's cloze task. A new trick we're calling <em>Language Modelling with Approximate Outputs</em> is used to apply the pre-training to smaller models. The pre-training outputs CNN and embedding weights that can be used in <code>spacy train</code>, using the new <code>-t2v</code> argument. <strong>NEW:</strong> Allow parser to do joint word segmentation and parsing. If you pass in data where the tokenizer over-segments, the parser now learns to merge the tokens. Make parser, tagger and NER faster, through better hyperparameters. Add simpler, GPU-friendly option to <code>TextCategorizer</code>, and allow setting <code>exclusive_classes</code> and <code>architecture</code> arguments on initialization. Add <code>EntityRecognizer.labels</code> property. Remove document length limit during training, by implementing faster Levenshtein alignment. Use Thinc v7.0, which defaults to single-thread with fast <code>blis</code> kernel for matrix multiplication. Parallelisation should be performed at the task level, e.g. by running more containers. Models & Language Data <strong>NEW:</strong> 2-3 times faster tokenization across all languages at the same accuracy! <strong>NEW:</strong> Small accuracy improvements for parsing, tagging and NER for 6+ languages. <strong>NEW:</strong> The English and German models are now available under the MIT license. <strong>NEW:</strong> Statistical models for Greek. <strong>NEW:</strong> Alpha support for Tamil, Ukrainian and Kannada, and base language classes for Afrikaans, Bulgarian, Czech, Icelandic, Lithuanian, Latvian, Slovak, Slovenian and Albanian. Improve loading time of <code>French</code> by ~30%. CLI <strong>NEW:</strong> <code>pretrain</code> command for ULMFit/BERT/Elmo-like pretraining (see #2931). <strong>NEW:</strong> New <code>ud-train</code> command, to train and evaluate using the CoNLL 2017 shared task data. Check if model is already installed before downloading it via <code>spacy download</code>. Pass additional arguments of <code>download</code> command to <code>pip</code> to customise installation. Improve <code>train</code> command by letting <code>GoldCorpus</code> stream data, instead of loading into memory. Improve <code>init-model</code> command, including support for lexical attributes and word-vectors, using a variety of formats. This replaces the <code>spacy vocab</code> command, which is now deprecated. Add support for multi-task objectives to <code>train</code> command. Add support for data-augmentation to <code>train</code> command. Other <strong>NEW:</strong> Enhanced pattern API for rule-based <code>Matcher</code> (see #1971). <strong>NEW:</strong> <code>Doc.retokenize</code> context manager for merging and splitting tokens more efficiently. <strong>NEW:</strong> Add support for custom pipeline component factories via entry points (#2348). <strong>NEW:</strong> Implement fastText vectors with subword features. <strong>NEW:</strong> Built-in rule-based NER component to add entities based on match patterns (see #2513). <strong>NEW:</strong> Allow <code>PhraseMatcher</code> to match on token attributes other than <code>ORTH</code>, e.g. <code>LOWER</code> (for case-insensitive matching) or even <code>POS</code> or <code>TAG</code>. <strong>NEW:</strong> Replace <code>ujson</code>, <code>msgpack</code>, <code>msgpack-numpy</code>, <code>pickle</code>, <code>cloudpickle</code> and <code>dill</code> with our own package <code>srsly</code> to centralise dependencies and allow binary wheels. <strong>NEW:</strong> <code>Doc.to_json()</code> method which outputs data in spaCy's training format. This will be the only place where the format is hard-coded (see #2932). <strong>NEW:</strong> Built-in <code>EntityRuler</code> component to make it easier to build rule-based NER and combinations of statistical and rule-based systems. <strong>NEW:</strong> <code>gold.spans_from_biluo_tags</code> helper that returns <code>Span</code> objects, e.g. to overwrite the <code>doc.ents</code>. Add warnings if <code>.similarity</code> method is called with empty vectors or without word vectors. Improve rule-based <code>Matcher</code> and add <code>return_matches</code> keyword argument to <code>Matcher.pipe</code> to yield <code>(doc, matches)</code> tuples instead of only <code>Doc</code> objects, and <code>as_tuples</code> to add context to the <code>Doc</code> objects. Make stop words via <code>Token.is_stop</code> and <code>Lexeme.is_stop</code> case-insensitive. Accept <code>"TEXT"</code> as an alternative to <code>"ORTH"</code> in <code>Matcher</code> patterns. Use <code>black</code> for auto-formatting <code>.py</code> source and optimse codebase using <code>flake8</code>. You can now run <code>flake8 spacy</code> and it should return no errors or warnings. See <code>CONTRIBUTING.md</code> for details. 🔴 Bug fixes Fix issue #1487: Add <code>Doc.retokenize()</code> context manager. Fix issue #1537: Make <code>Span.as_doc</code> return a copy, not a view. Fix issue #1574: Make sure stop words are available in medium and large English models. Fix issue #1585: Prevent parser from predicting unseen classes. Fix issue #1642: Replace <code>regex</code> with <code>re</code> and speed up tokenization. Fix issue #1665: Correct typos in symbol <code>Animacy_inan</code> and add <code>Animacy_nhum</code>. Fix issue #1748, #1798, #2756, #2934: Add simpler GPU-friendly option to <code>TextCategorizer</code>. Fix issue #1773: Prevent tokenizer exceptions from setting <code>POS</code> but not <code>TAG</code>. Fix issue #1782, #2343: Fix training on GPU. Fix issue #1816: Allow custom <code>Language</code> subclasses via entry points. Fix issue #1865: Correct licensing of <code>it_core_news_sm</code> model. Fix issue #1889: Make stop words case-insensitive. Fix issue #1903: Add <code>relcl</code> dependency label to symbols. Fix issue #1963: Resize <code>Doc.tensor</code> when merging spans. Fix issue #1971: Update <code>Matcher</code> engine to support regex, extension attributes and rich comparison. Fix issue #2014: Make <code>Token.pos_</code> writeable. Fix issue #2329: Correct <code>TextCategorizer</code> and <code>GoldParse</code> API docs. Fix issue #2369: Respect pre-defined warning filters. Fix issue #2390: Support setting lexical attributes during retokenization. Fix issue #2396: Fix <code>Doc.get_lca_matrix</code>. Fix issue #2464, #3009: Fix behaviour of <code>Matcher</code>'s <code>?</code> quantifier. Fix issue #2482: Fix serialization when parser model is empty. Fix issue #2644: Add table explaining training metrics to docs. Fix issue #2648: Fix <code>KeyError</code> in <code>Vectors.most_similar</code>. Fix issue #2671, #2675: Fix incorrect match ID on some patterns. Fix issue #2693: Only use <code>'sentencizer'</code> as built-in sentence boundary component name. Fix issue #2728: Fix HTML escaping in <code>displacy</code> NER visualization and correct API docs. Fix issue #2754, #3028: Make <code>NORM</code> a <code>Token</code> attribute instead of a <code>Lexeme</code> attribute to allow setting context-specific norms in tokenizer exceptions. Fix issue #2769: Fix issue that'd cause segmentation fault when calling <code>EntityRecognizer.add_label</code>. Fix issue #2772: Fix bug in sentence starts for non-projective parses. Fix issue #2779: Fix handling of pre-set entities. Fix issue #2782: Make <code>like_num</code> work with prefixed numbers. Fix issue #2833: Raise better error if <code>Token</code> or <code>Span</code> are pickled. Fix issue #2838: Add <code>Retokenizer.split</code> method to split one token into several. Fix issue #2870: Make it illegal for the entity recognizer to predict whitespace tokens as <code>B</code>, <code>L</code> or <code>U</code>. Fix issue #2871: Fix vectors for reserved words. Fix issue #2901: Fix issue with first call of <code>nlp</code> in Japanese (MeCab). Fix issue #2924: Make IDs of displaCy arcs more unique to avoid clashes. Fix issue #3012: Fix clobber of <code>Doc.is_tagged</code> in <code>Doc.from_array</code>. Fix issue #3027: Allow <code>Span</code> to take unicode value for <code>label</code> argument. Fix issue #3048: Raise better errors for uninitialized pipeline components. Fix issue #3064: Allow single string attributes in <code>Doc.to_array</code>. Fix issue #3093, #3067: Set <code>vectors.name</code> correctly when exporting model via CLI. Fix issue #3112: Make sure entity types are added correctly on GPU. Fix issue #3122: Correct docs of <code>Token.subtree</code> and <code>Span.subtree</code>. Fix issue #3128: Improve error handling in converters. Fix issue #3248: Fix <code>PhraseMatcher</code> pickling and make <code>__len__</code> consistent. Fix issue #3277: Add en/em dash to tokenizer prefixes and suffixes. Fix serialization of custom tokenizer if not all functions are defined. Fix bugs in beam-search training objective. Fix problems with model pickling. ⚠️ Backwards incompatibilities This version of spaCy requires downloading <strong>new models</strong>. You can use the <code>spacy validate</code> command to find out which models need upda
Implicit discourse relation classification is one of the most challenging and important tasks in discourse parsing, due to the lack of connectives as strong linguistic cues. A principle bottleneck to further improvement is the shortage of training data (ca. 18k instances in the Penn Discourse Treebank (PDTB)). Shi et al. (2017) proposed to acquire additional data by exploiting connectives in translation: human translators mark discourse relations which are implicit in the source language explicitly in the translation. Using back-translations of such explicitated connectives improves discourse relation parsing performance. This paper addresses the open question of whether the choice of the translation language matters, and whether multiple translations into different languages can be effectively used to improve the quality of the additional data.
The Malayalam language possesses characteristics like morphological richness and agglutination, which makes the processing of language a tiresome task. The resources like treebank allow the syntactic-semantic analysis of Malayalam. In this work, the validation and optimization of Malayalam Treebank Data_ Iiithwere done with various parsing settings. The improvement in the values of qualitative analysis metrics was the main result of the optimization experiments. 10-cross-validation of the data was also performed for the analysis of the attainment of models generated. The results were found to be promising.
Abstract How might language have evolved and which (types of) norms, if any, might have played a role in shaping it? This chapter addresses these two questions by first exploring differences between human language and animal communication systems; the difference between natural signs, signals, and non-natural signs (symbols) will be elaborated. The author claims that normativity enters the picture only at the level of symbols. The remainder of the chapter is devoted to the question of what kinds of norms might have played a role in the evolution of symbolic communication. The author argues, firstly, that a certain level of cooperation is needed if non-natural signs are to be interpretable at all; secondly, a type of prudential norm emerges as signs acquire stable meaning; thirdly, interpretation of implicit communication is governed by pragmatic norms, too.
Insight has been investigated under the assumption that participants solve insight problems with insight processes and/or experiences. A recent trend has involved presenting participants with the solution and analysing the resultant experience as if insight has taken place. We examined self-reports of the aha experience, a defining aspect of insight, before and after feedback, along with additional affective components of insight (e.g., pleasure, surprise, impasse). Classic insight problems, compound remote associates, and non-insight problems were randomly interleaved and presented to participants. Solution feedback increased ratings of aha experience in both insight and non-insight problems, with this result being driven by responses to solutions that were initially incorrectly generated. Ratings of aha for correctly generated solutions decreased after the correct solution was presented. These findings have implications for insight research paradigms as well as informing teaching methods.
This paper presents a syntactic treebank for spoken Naija, an English pidgincreole, which is rapidly spreading across Nigeria. The syntactic annotation is developed in the Surface-Syntactic Universal Dependency annotation scheme (SUD) We present the workflow of the treebank development for this under-resourced language. A crucial step in the syntactic analysis of a spoken language consists in manually adding a markup onto the transcription, indicating the segmentation into major syntactic units and their internal structure. We show that this so-called "macrosyntactic" markup improves parsing results. We also study some iconic syntactic phenomena that clearly distinguish Naija from English.
In this paper we describe the early stage application of the Universal Dependencies to an Italian corpus from social media developed for shared tasks related to irony and stance detection.The development of this novel resource (TWITTIR Ò-UD) serves a twofold goal: it enriches the scenario of treebanks for social media and for Italian, and it paves the way for a more reliable extraction of a larger variety of morphological and syntactic features to be used by sentiment analysis tools.On the one hand, social media texts are especially hard to parse and the limited amount of resources for training and testing NLP tools further damages the situation.On the other hand, we thought that adding the Universal Dependencies format to the fine-grained annotation for irony, that was previously applied on TWITTIR Ò, might meaningfully help in the investigation of possible relationships between syntax and semantics of the uses of figurative language, irony in particular.
Abstract How to classify short texts effectively remains an important question in computational stylometry. This study presents the results of an experiment involving authorship attribution of ancient Greek texts. These texts were chosen to explore the effectiveness of digital methods as a supplement to the author’s work on text classification based on traditional stylometry. Here it is crucial to avoid confounding effects of shared topic, etc. Therefore, this study attempts to identify authorship using only morpho-syntactic data without regard to specific vocabulary items. The data are taken from the dependency annotations published in the Ancient Greek and Latin Dependency Treebank. The independent variables for classification are combinations generated from the dependency label and the morphology of each word in the corpus and its dependency parent. To avoid the effects of the combinatorial explosion, only the most frequent combinations are retained as input features. The authorship classification (with thirteen classes) is done with standard algorithms—logistic regression and support vector classification. During classification, the corpus is partitioned into increasingly smaller ‘texts’. To explore and control for the possible confounding effects of, e.g. different genre and annotator, three corpora were tested: a mixed corpus of several genres of both prose and verse, a corpus of prose including oratory, history, and essay, and a corpus restricted to narrative history. Results are surprisingly good as compared to those previously published. Accuracy for fifty-word inputs is 84.2–89.6%. Thus, this approach may prove an important addition to the prevailing methods for small text classification.
Introduction. The problems of linguistic norm as an important component of the modern linguistic and cultural space as an integral attribute of discursive practices have been considered. The key method of a person’s adaptation to specific needs of the communicative space is to go beyond the codified norm – to ignore it or, conversely, to pay attention to every subtle nuance of its criteria. Aim and tasks. The publication aims to analyze the social and cultural factors of transformation of the general framework of the speech culture in the modern informational era, in particular, those of the linguistic norm. Research methods. The methodological tools consist of the sociocultural approach, methodological principles of systemacity and completeness, as well as the method of cultural-semantic analysis. Research results. In the information society, the linguistic norm has an ambivalent character: on the one hand, it is perceived as a natural manifestation of the adaptive potential of the language in the new conditions of information exchange, on the other – as a factor posing a potential threat to the formal discourse at all levels. Discussion. The results of the analysis of the linguistic norm in the conditions of the informational era are complementary to the conceptual developments of researchers in the context of modern globalization transformations. Conclusion. The key prerequisites for active blurring of the linguistic norm limits have been established, resulting in a paradoxical situation, where the lack of norm is considered to be a norm. In the process of transformation of the linguistic norm, a special attention is paid to the phenomenon of intensive convergence of the spoken and written speech, which is clearly traced in the analysis of virtual communication. The core trends of the latter are excessive expressiveness, a tendency to conciseness, simplification and, at the same time, a more serious approach to the word choice, which is clearly shown in choosing keywords for data search.