Towards Learning a Universal Non-Semantic Representation of Speech

The purpose of media programming is to inform and to entertain. Some programs can be classified as informative, others as entertainment, and others as both. The relationship between language and the media is particularly interesting in the Arabic context due to the diglossic nature of its speech communities. The media, particularly the broadcasting media, has a unique role to play in negotiating the relationship between the two varieties: spoken everyday language of communication and a literary variety. This chapter discusses the position that the broadcasting media represent 'hybrid' contexts, and focuses on the interview as a genre whose purpose is to inform and entertain, and whose context is appropriate for either or both varieties of Arabic. It examines two interviews aired over the Egyptian Satellite Channel. Three major styles are identified, corresponding roughly to Labov's reading, careful, and casual styles. Finally, it discusses 'hybridity' in relation to 'code-switching'. Keywords: Arabic; hybridity; language of communication; media programming; speech communities

This paper discusses a common reality in many cases of multilingualism: heritage speakers, or unbalanced bilinguals, simultaneous or sequential, who shifted early in childhood from one language (their heritage language) to their dominant language (the language of their speech community). To demonstrate the relevance of heritage linguistics to the study of linguistic competence more broadly defined, we present a series of case studies on heritage linguistics, documenting some of the deficits and abilities typical of heritage speakers, together with the broader theoretical questions they inform. We consider the reorganization of morphosyntactic feature systems, the reanalysis of atypical argument structure, the attrition of the syntax of relativization, and the simplification of scope interpretations; these phenomena implicate diverging trajectories and outcomes in the development of heritage speakers. The case studies also have practical and methodological implications for the study of multilingualism. We conclude by discussing more general concepts central to linguistic inquiry, in particular, complexity and native speaker competence.

The present thesis is the first thorough grammatical description of Romeyka, i.e., the Muslim variety of Pontic Greek still spoken in the indigenous setting in Trabzon province in north-eastern Turkey. While the language of the Christian speech communities of Pontus has been featured in earlier research during the last hundred years – both in Pontus prior to the population exchange in 1923, and after that in Greece –, less research is available on the variety of Pontic Greek as spoken by Muslims in the area at present. As an unrecognized minority language in a Turkish-dominant majority society and due to ambivalent attitudes and identificational links of the speakers towards their language, Romeyka is currently undergoing language shift to Turkish. While the language has been in contact with Turkish for several hundred years – which arguably resulted in a number of lexical and structural contact influences albeit without leading to a mixed acrolect –, the influence of Turkish on domains where traditionally Romeyka used to be spoken has grown following labour migration from the Black Sea area to larger cities in western Turkey since the 1960s posing now – dependent upon the speech community – a threat to the vitality of Romeyka. On the other hand, Pontic Greek – and in particular the Muslim variety – is known to be a conservative representative of Asia Minor Greek preserving archaisms dating back via Hellenistic and Medieval Greek to Ancient Greek, among them so notorious grammatical features as the Romeyka infinitive, but also complex negators and pronominal forms. It is the aim of the present thesis to describe the synchronic grammar of Romeyka as currently spoken especially by elderly speakers in the Of valley, whose variety can be considered the most archaic. As a basis for profound grammatical research – and as a contribution towards language description and documentation at the same time –, a spoken language corpus has been compiled from the variety Romeyka of Of as spoken in Çaykara based on recordings of naturalistic oral texts produced by ten speakers mainly during fieldwork in the area in 2019. In order to contribute to further scholarly research, the corpus has been transcribed, translated and fully grammatically annotated and will be made publicly available upon the publication of the thesis. In the light of the ongoing language shift in Romeyka, the present thesis focuses on the actual use of the language by bilingual – and most often Turkish dominant – speakers. In order to disentangle the factors that led to the present grammatical shape of Romeyka, three research questions targeting potential causes for the grammatical developments have been considered: (I.) How strong is the Turkish influence on the grammar of Romeyka? (II.) Does the setting of Romeyka as a minority language in a Turkish-dominant society, namely the process of language shift, play a role in grammatical change? (III.) How persistent are the inherited Greek features and what is the role of language-internal changes? However, since the primary aim of the present thesis is the description and presentation of linguistic structures forming potential candidates for an ongoing language change compared to earlier varieties of Greek and to closely related varieties of Asia Minor Greek, the theoretical questions can only tentatively be answered at the present stage of research. It can be nevertheless stated that Romeyka seems to have preserved its Greek core with a Greek basic lexicon, albeit with considerable lexical influences from Turkish in both content and functional word classes, but probably to a lesser extent compared to its close relative Cappadocian. Furthermore, Romeyka appears to preserve indeed a number of archaic features as well as certain particularities of Pontic such as a complex system of spatial deixis. On the other hand, there are considerable Turkish influences on the structural domain such as for example in progressives, verb serialization, adjunct and relative clauses and clause combination, that have evolved under long-time contact and appear consistently across different speakers. In addition, certain contact-induced structural features especially at the syntactic domain have been found to be sensitive to the bilingual profile of the individual speaker, i.e., among others, the dominant language, namely the borrowing of the Turkish question particle mI, OV word orders in declarative clauses, and an increase in non-finite complementation strategies. This finding led to the assumption that – apart from long-term contact-induced changes which seem not to impair the grammatical integrity of Romeyka – language shift is a matter of individual multilingual speaker profiles with different degrees of Turkish dominance in both language competence and use. Finally, it is argued that the specific communicative setting in which Romeyka is spoken, and which determines the communicative practice of the speakers, is decisive to set the pragmatic frame for the language(s) used: in Turkish-dominant settings, speakers tend to be more inclined to apply multiple forms of code-switching and show stronger influences of Turkish structures.

Code-switching has become a common social phenomenon governed by social conversational needs. Central to the use of code-switching is the way in which social norms, which are also called rights and obligations, are attributed to speakers and listeners of certain social categories. Studies on code-switching reveal that its occurrence is indicative of group membership in speech communities where the use of more than one language in communication is the norm. Code-switching is looked upon as the practice of altering elements of language in order to contextualise talk in interaction. The contextualisation referred to may relate to discourse practices or make information beyond the current exchange relevant, including knowledge of society. This article seeks to investigate and critically analyse the use of code-switching by Nhlapho (1996) in the novel, Imbali YemaNgcamane (The flower of the Ngcamanes). It also seeks to explore whether the use of code-switching benefits or has adverse effects on the speakers of the Siswati language.

This article reports on Bajjika, a minority language of the Bihari group of the Indo-Aryan language family. Bajjika is spoken by millions of speakers in the northern part of the north Indian state of Bihar and more than 200,000 speakers in Nepal. Although Bajjika is spoken by a significant segment of the population of Bihar, it has no institutional or educational role in the community or government. Little is known about this language in terms of demographic and ethno-linguistic details, not only to the world but also within India beyond the speech community. This article provides details about the speech community of Bajjika, its language status, functional role in the community and the revival efforts made by exponents of the language. Finally, suggestions for the maintenance of this language are made.

Machine learning (ML) has emerged as a critical enabling tool in the sciences and industry in recent years. Today’s machine learning algorithms can achieve outstanding performance on an expanding variety of complex tasks–thanks to advancements in technique, the availability of enormous databases, and improved computing power. Deep learning models are at the forefront of this advancement. However, because of their nested nonlinear structure, these strong models are termed as “black boxes,” as they provide no information about how they arrive at their conclusions. Such a lack of transparencies may be unacceptable in many applications, such as the medical domain. A lot of emphasis has recently been paid to the development of methods for visualizing, explaining, and interpreting deep learning models. The situation is substantially different in safety-critical applications. The lack of transparency of machine learning techniques may be limiting or even disqualifying issue in this case. Significantly, when single bad decisions can endanger human life and health (e.g., autonomous driving, medical domain) or result in significant monetary losses (e.g., algorithmic trading), depending on an unintelligible data-driven system may not be an option. This lack of transparency is one reason why machine learning in sectors like health is more cautious than in the consumer, e-commerce, or entertainment industries. Explainability is the term introduced in the preceding years. The AI model’s black box nature will become explainable with these frameworks. Especially in the medical domain, diagnosing a particular disease through AI techniques would be less adapted for commercial use. These models’ explainable natures will help them commercially in diagnosis decisions in the medical field. This paper explores the different frameworks for the explainability of AI models in the medical field. The available frameworks are compared with other parameters, and their suitability for medical fields is also discussed.

Language is a tricky thing.On the one hand, it's a matter of convention, of speech communities and their shared communicative systems.On the other hand, it's a cognitive activity of individual speakers, and no two people talk alike.How do these two levels interact?What is the relationship between the macro speech community and its language, and the micro level of individual speakers and their language?For a long time, linguists have shunned that question and have concentrated on the conventionalized linguistic system, emphasizing the goal to achieve greatest possible generalizations with their findings.Notable exceptions are mostly found in sociolinguistic studies such as Bergs ( 2005), Johnstone (1996( ), or Milroy (1992)).Pentrel ( 2017) was among the first to highlight and investigate the cognitive mechanisms of individual language use in historyin that case the language of Samuel Pepys.The Mind Bending Grammars project 1 (initiated by Peter Petr) was probably the first large scale research project focusing on both the sociolinguistic and cognitive underpinnings of individual and communal language use, and Lynn Anthonissen's study Individuality in language change is the first book-length culminating point for these developments.Anthonissen singles out two particular case studies of constructional change in English that illustrate the interconnectivity of individual and community grammars.These two are the prepositional passive construction (The issue was frowned upon) on the one hand, and the nominative and infinitive construction (NCI; She is said to be a hero) on the other.She investigates these constructions and their changes on the basis of the Early Modern Multiloquent Authors corpus (EMMA).The EMMA corpus comprises about 90 million words from about 50 different, but interconnected seventeenth-century authors across five generations.This book is divided into two main parts and a concluding section.Part II (Chapters 2-4) focuses on the two special passive constructions at hand and their historical developments.Part III (Chapters 5-8) then concentrates on the relationship and interplay of individual speakers and the language community with regard to the findings of part II.Part IV (Chapters 9 and 10) offers some concluding remarks.

The development of terminology features heavily in language planning, and here the differences between planned and ethnic languages are much less pronounced. This is especially the case in languages with smaller numbers of speakers, or in indigenous and endangered languages such as Te Reo Māori of Aotearoa New Zealand that rely on language planning for their survival, and where conscious terminology planning is therefore commonplace. The present article compares the terminological principles that are applied in the creation of new terms in Te Reo Māori and the planned language Esperanto. Different preferences for endogenous versus exogenous ways of developing new words generate conflict in both language communities as they adapt to the demands of functioning in modern and international arenas. Long-term success in terminological planning can only be achieved by more comprehensive application of principles from terminological science to maximize the adequacy of the generated terms and their acceptance within the speech communities.

The globalization of English has taken the language to many parts of the world. Rather than being a language owned and used mainly by speech communities that claim it as their mother tongue, it is increasingly being used as a language for international and intercultural communication. The notion of “English as an International Language” is a proactive response to this situation and is reconceptualizing English in relevant theoretical and practical domains.

To understand language choice and development in the process of industrialization it is essential to study how the children of workers (the second generation) in a rising industrial community choose their language/dialect for daily communication. The coexistence of three speech communities: dialect, dialect-and-Putonghua mixed, and Putonghua, in three urban districts of Luoyang City, Henan Province, since the 1950s is representative of language development during China’s industrialization. Based on a large-scale survey, this article compares language and dialect use in these three urban districts and some special danwei (work-unit) language islands in Luoyang. This study shows why and how the second generation in a rising industrial community have chosen or not chosen Putonghua as their major language for daily communication. It concludes with a discussion on the characters and underlining principles of language choice in the course of industrialization under the planned economy in China.


 
 
 Food-market speech is an under-researched area of third-wave variationist sociolinguistic studies. This study addresses the gap by exploring food-market speech styles and hawker personae. Combining descriptive auditory analysis and online questionnaire data, I demonstrate that situated discursive practices of prosodic variables construct both persuasive and aggressive speech styles, and they are stereotypically associated with female and male hawker personae. Furthermore, this paper also explores the ideological construal of hawking as authentic market-ness, further revealing the semiotic saliency and social significance of food-market hawking as not only the language of a speech community but the language of market.
 
 

A common approach to language awareness is through the contrast between folk and scientific knowledge, the former usually getting brief attention and being assigned little value. This paper argues that the folk awareness of language may be characterised in several different ways, having to do with a wide variety of linguistic characterisations (e.g. overt ‘availability’, ‘degree of accuracy’, ability to ‘control’ language varieties, and the level of specificity of folk knowledge). Only the second of these has to do directly with the folk versus linguist dichotomy, although cognitive constraints on what non‐linguists ‘usually’ know are important factors. More important, however, at least for speakers of American English, are the constraints imposed by notions of language ‘correctness’. In addition, folk linguistics is shown here to be a dynamic area of study, perhaps best uncovered by the analysis of ongoing discourses about language in speech communities. Finally, the value of folk linguistic awareness for both its scientific merit and its importance to applied concerns is briefly discussed.

This essay argues that the pervasive twentieth century understanding of meaning — a sign stands for an object — is incorrect. In its place, it offers the following definition, which is framed not in terms of a single relation (of standing for), but in terms of a relation (of correspondence) between two relations (of standing for): a sign stands for its object on the one hand, and its interpretant on the other, in such a way as to make the interpretant stand in relation to the object corresponding to it own relation to the object. Using this definition, it reanalyzes key concepts and foundational arguments from linguistics so far as they relate to anthropology and psychology. Such terms include: concept, intentional state, motivation, ground, iconicity, speech community, norm, performativity, joint-attention, embodiment, intersubjectivity, agency, role, functionalism, pragmatics, social construction, realism, and natural language.

There seems to be nothing remarkable about the interaction between two interlocutors who have never been in contact with each other. These persons are able to understand themselves in contact situations because most times, a common language of communication is known that can sustain the exchange for the time necessary. However, when such exchange is between individuals with some level of contact or familiarity, the concept of speech community comes into play. The concept is useful but may be problematic at times and one cannot avoid applying this idea when trying to make sense of the process that takes place in the conversation, specifically a causal conversation. The aim of this sociolinguistic study is to explain how individuals are able to build social history, construct interactional talk, maintain relations with each other and reinforce solidarity from a two hour audio recorded conversation (ARC) between an ethnic Indian and a Nigerian in Marylebone, London using interactional socio-linguistic and conversation analytic. By doing so, the concept of a speech community as well as how a group can be identified as being members of a community is understood. A particular focus is paid to such linguistic features as the register of conversation, turn taking, discourse variation, phonological variation and grammatical variation characteristic of London, Nigerian and Indian English observed in the speech of the participants and how these features function to build and maintain relations.
 Keywords: Speech community, Casual conversation, Linguistic features, Sociolinguistics, ARC

Summary Language death is a phenomenon with symptoms related to demeaning vocabulary count and depletion of domains of language use along with the simplicity of linguistic structures. It commences by exhibiting traits of a declining number of fluent speakers, dwindling attitude of the speakers with regard to their heritage language, language shift, lack of inter-generational language transfer accompanied by a feeling that heritage language is inferior to outside languages, and Gulgulia exhibits every trait of such a dying language. It has become a waning language that is very close to its permanent extinction. The present study elucidates the ethnolinguistic vitality of Gulgulia tested through chosen sociolinguistic parameters which were found suitable to Gulgulia’s scenario. It also explores the linguistic situation of the Gulgulian community in Dhanbad, the community members’ language use in their homes and in their miscellaneous interethnic interactions and examines what relation prevails between the community’s language preference and their vitality. It was found that speaker variables, such as age, gender, and language competence governed the speaker’s attitude toward the heritage language. The location of the speech community is also a regulating factor in determining the inclination for preservation or attrition of the native language. The analysis of the speech behavior in the Gulgulian community confirms the loss of major genres such as the art of narration. Out of all the genres of language use, only two are surviving, which is alarming.