Monolingual Speech Research Articles

“…because dem Computer brauchst' es ja nicht zeigen.”*: because + German main clause word order

This paper aims to explain firstly, different frequency distributions of the causal subordinators because and weil in monolingual and bilingual data sets and secondly, verb second, that is, German main clause word order, in clauses introduced by these two conjunctions. Quantitative and quantitative methods will be used. The quantitative study of the mixed German/English data reveals almost twice as many because than weil in a corpus that otherwise has more German clause linkers than English ones. The qualitative analysis exposes verb second rather than verb final word order in predominantly German clauses introduced by because and weil. I will show first, that we understand enough about the interdependence between frequency distributions of word order patterns in monolingual speech and properties of the monolingual grammars to handle frequency in bilingual data. We can utilize frequency figures from monolingual and mixed corpora and combine them with syntactic and pragmatic knowledge to explain the interdependence between grammatical patters and frequency distributions in bilingual data. Secondly, I will demonstrate that it is the availability of context in language corpora that makes it feasible to examine the hypothesis that foreign conjunctions do not trigger verb-final word order in German clauses because they are used in functions that require main clause order.

Acquisition by Processing Theory: A theory of everything?

Truscott and Sharwood Smith (henceforth T&SS) propose a novel theory of language acquisition, ACQUISITION BY PROCESSING THEORY (APT), designed to account for both first and second language acquisition, monolingual and bilingual speech perception and parsing, and speech production. This is a tall order. Like any theoretically ambitious enterprise, the APT shares certain properties with much that has gone before. Like the Competition Model (CM; MacWhinney, 1987, 1997; MacWhinney and Bates, 1989, inter alia) and other associative network connectionist learning models, the APT eschews a Language Acquisition Device (LAD) by treating acquisition as the strengthening of levels of representation activation. A parser can produce multiple representations of a parse string in parallel, which then ‘compete’ as analyses for an input string. Unlike the CM, however, the APT is not motivated by a solid program of empirical studies in language acquisition or cross-language processing. Nor does it strike me as theoretically coherent, for the APT, unlike the CM, assumes that knowledge of language involves knowledge of grammatical structure and that the parser makes deterministic use of Universal Grammar in the form of a Minimalist grammar. The determinism is important here; the claim to eliminate LAD hinges on it.

How code-switching mediates politeness

In this study we explore the three-way link between three topics which are usually studied independently in sociolinguistics, namely code-switching (CS), gender and politeness. Whilst steering clear of simplistic correlations between language and gender, we show how women in this community exploit CS in carrying out certain direct speech acts traditionally viewed as unfeminine. Conversational analysts often describe CS as a site for constructing meaning through juxtaposition, the direction of the switch being less important than the fact of the contrast. Nevertheless, an ethnographic study such as this shows the potential importance of associations between each language and relevant shared meanings. Switches into Greek are employed to evoke shared cultural connotations and by extension to create humour, to create a bond/show sympathy to the interlocutor, and to legitimize directness in giving orders or making requests, Greek being a language in which such directness is markedly greater than in English. Thus, through their use of CS, women in particular enact various politeness strategies, both positive and negative. This study confirms the value of studying both politeness and gender not only cross-culturally, but specifically in bi/plurilingual contexts, as a study of CS allows strategies and constraints which could pass unobserved in monolingual speech to be clearly highlighted.

A Sociophilological Study of Late Latin (review)

254 Reviews Parergon 20.2 (2003) Wright, Roger, A Sociophilological Study of Late Latin (Utrecht Studies in Medieval Literacy 10), Turnhout, Brepols, 2002; cloth; pp. viii, 389; RRP EUR65.00; ISBN 2503513387. In 1982, Roger Wright published an important volume on the complex transition from spoken Latin to the Romance languages of southern Europe (Late Latin and Early Romance in Spain and Carolingian France). In it, he argued that the former western provinces of the Roman empire continued to be a monolingual speech community long after the collapse of political unity. More-or-less standardised written Latin remained widely used and comprehensible, with much the same relationship to local variations of spoken Latin as modern written English has to its myriad of spoken accents, dialects, and creoles today. To Wright, the crucial period of transition from monolingualism (with variants) to a Latin/Romance dichotomy was the Carolingian ‘Renaissance’, and the prime agents of change were the Anglo-Saxon scholars at Charlemagne’s court. Alcuin and his ilk introduced a hyper-corrected form of spoken Latin which was quite at odds with Latin as it was still actually spoken on the Continent, not least in its practitioners’ insistence on full pronunciation of all written letters. Spread via the Carolingian educational reforms, this hyper-correction drove a wedge between the written and the spoken language. The latter, by the thirteenth century at the latest, had become a variety of vernaculars alienated from the written language; while written Latin had assumed that position we all know, of a ‘dead’ foreign language, universally available but possessed by none. The present volume further develops major aspects of Wright’s thesis. The book is a collection of articles (a fact unannounced on the cover blurb or advertising), most originally published within a few years either side of the millennium. The papers are mildly reworked (including translation of several from three Romance languages) in order to provide coherence, but the author’s clearly demarcated research interests provide a high degree of unity. The book is the third release in a potentially very interesting series on medieval literacy. The chapters are divided into six main themes: longitudinal overviews of relations between late and medieval Latin and the Romance languages; four sections on periods and regions which were either key points of linguistic development (Late Antiquity; the Carolingian period; and twelfth/thirteenth century Spain) or instructive isolates (tenth/eleventh century Italy and Spain); and methodological studies in ‘sociophilology.’ The latter term is the author’s coinage. It refers to a marriage between sociolinguistics (examining the social Reviews 255 Parergon 20.2 (2003) agenda of linguistic variation), historical linguistics (concerned with spoken language development in pre-literate groups), and the traditional, technical philology needed to be able to deal with such remote data. Wright is a vigorous and precise speaker, and his papers reflect the same forceful orderliness. The titles of his compact chapters declare a problem-solving directness: ‘How Latin Came to Be a Foreign Language for All’, ‘Why the Romance Languages Are Not All the Same’, ‘Why Judeo-Spanish Was Called “Ladino”’, ‘What Actually Changes During a “Sound Change”?’As subsidiaries to his main theme, Wright introduces a range of fascinating insights into cultural change and continuity in the early Middle Ages. One example which recurs several times throughout the book is Ladino, the language of the Mozarabic Christian community under Islamic rule in Córdoba, a fossil of what might be called Wright’s ‘prelapsarian ’common Latin, unaffected by Carolingian self-consciousness of ‘correct’ Latin and ‘non-Latinate’ Romance. But while the spoken form of Ladino continued to be used, by the tenth century its written counterpart had fallen into disuse. The Ladino-speaking community bowed to convenience, abandoned classicising Latin as a written mode, and adopted as their written language what was by now its second spoken tongue, Arabic (itself a highly standardised language). The papers on Spain in the central Middle Ages (section E) provide the most detailed arguments for the wave of conscious change in the introduction of Carolingian-style standardisation of Latin, relating changes in particular areas to the episcopates of individual bishops. There is significant overlap with manuscript studies in this section...

Parallel patterns? A comparison of monolingual speech and bilingual codeswitching discourse

The extensive work done on the structure of monolingual discourse is now paralleled by a strong tradition of studies of the conversational functions of bilingual codeswitching (Gumperz, 1982; Myers-Scotton, 1993a; Auer, 1998a). So far, however, no direct comparisons have been made between the two. In this paper we compare the way in which four common conversational functions are realised (a) monolingually and (b) through codeswitching by members of a Punjabi and English-speaking network in London. The samples are thus ideally matched - the same speakers in the same context - and we establish that codeswitching may be used in two ways within these conversations. On the one hand it may take the place of monolingual ways of marking significant moves in the conversation (e.g. emphasis, change in voice quality), or add itself to these to reinforce the effect. On the other hand it can be used as a further dimension to the monolingual means which are available, allowing the speakers to introduce structural contrasts, manage the conversational ‘floor’, or highlight the different connotations of each variety as a counterpoint to the referential meaning of their utterance.

Testing the 4-M model: An introduction

The goal of this special issue is to report on new directions in our research that began with codeswitching, but increasingly includes other types of language contact phenomena. The issue consists of articles by the junior members of our research team and reports on their analyses and explanations for a variety of outcomes in bilingual production. Their approaches rely on the principles laid down in Duelling Languages (Myers-Scotton, 1993 [1997]). That is, their research looks for answers at an abstract level. Just like monolingual speech, bilingual speech is not best explainable only in terms of surface configurations. Duelling Languages presents an innovative model of structural constraints on codeswitching to explain the differential contributions of the participating languages in codeswitching. These constraints refer to the Matrix Language (ML) versus Embedded Language (EL) opposition and the content versus system morpheme opposition. Using these two oppositions, the model accounts for the bilingual structures observed in codeswitching data produced by proficient bilinguals. In addition to explicating the data in Duelling Languages, the Matrix Language Frame (MLF) model also accounts for the bilingual constituents from such diverse language pairs such as Turkish/Dutch (Backus, 1996), Chinese/English (Wei, 1999), Hungarian/English (Bolonyai, 1998), Arabic/English (Jake & Myers-Scotton, 1997), Spanish/English and Italian/Swiss German (Myers-Scotton & Jake, 1998b), and Mandinka, Wolof/English (Haust, 1995). The MLF model was based largely on what Myers-Scotton and Jake have since termed “classic codeswitching,” that is, codeswitching by speakers proficient enough in all participating varieties that they could engage in monolingual discourse in any of them. In Section 2 below, we offer readers

Initial Development Towards a Trilingual Speech Interface for Financial Information Inquiries

This paper describes our initial effort in developing a trilingual speech interface for financial information inquiries. Our foreign exchange inquiry system consists of: (i) monolingual and trilingual speech recognizers, which receive the user's spoken input in the form of microphone speech; (ii) a real-time data capture component which continuously updates a relational database from a financial data satellite feed; and (iii) a trilingual speech generation component, which generates English and Chinese text based on the raw financial data. The generated text is then transformed into spoken presentations. English text is processed by the FESTIVAL synthesizer system. Chinese text is sent to our syllable-based synthesizer, which employs a concatenative resequencing technique to produce spoken presentations in Putonghua or Cantonese. The speech interface is augmented with a visual display which aims to provide feedback to the user at all times during an interaction. Within the restricted scope of foreign exchange (FOREX), our recognition performance accuracies remain above 93%. Confusions across languages contributed significantly to our recognition errors, but most are confusions between the same currency/country names spoken in different languages. These errors are not detrimental with respect to data retrieval. Our concatenative re-sequencing technique reports the date, time and exchange rates of the input currency pair. A demonstration can be found at http://www.se.cuhk.edu.hk/hccl/demos/.

Duelling Languages: Grammatical Structure in Codeswitching

Myers-Scotton’s Direlliiig Laizgirqys is a book which deals with codeswitching (CS) from a sociolinguistic, but also, and predominantly, from a linguistic and a psycholinguistic point ofviewv. The book first came out in 1993, in hardback only, but has apparently been received so well that Oxford University Press decided to bring out a new paperback edition in 1997. In fact, the word “new” is slightly misleading, since the paperback edition is an exact copy of the first edition, but a twenty page aflenvord has been added to it. In the aflenvord MyersScotton clarifies some issues that had led to misunderstandings in the earlier edition and revises certain parts to bring them in line with her current thinking. The new book, then, consists of eight chapters and an aftenvord. In Chapter I, MyersScotton states the main goal of the book: “to present a model to account for the structures in intrasential CS” (p.5). In addition, Chapter I contains a detailed description of the database used for this study, the Nairobi corpus. IVe are given an interesting description of the position of Swahili and English in Kenya, and of the sociolinguistic background of the conversations from which the codeswitches were taken. Chapter I1 offers a critical review of earlier research on CS. One of the weaknesses pointed out is that in most early research single lexemes were excluded as instances of CS. The chapter is also critical of most early proposals for structural constraints, the main problem being that they have no theoretical basis and are therefore descriptive rather than explanatory. In Chapter 111, the background for the Matrix Language Frame (MLF) model is sketched. First, there is a discussion of several monolingual speech production models, notably Garrett’s (e.g., 1990) and Levelt’s (1989). Second, the distinction between the matrix language (ML) and the embedded language (EL) is introduced. A frequency-based criterion is suggested to operationalize the notion of the ML, which is defined as “the language of more morphemes in interaction types including intrasentential CS” (p. 68). Chapters IV and V constitute the t\vo most important chapters of the book since they present the MLF model and the predictions following from it. The MLF model is psycholinguistic in that it adopts some of the ideas from current work on speech production, such as the distinction between two kinds of morphemes: System morphemes and content

Language Contact in Japan: A Sociolinguistic History

The Japanese are often characterized as exclusive and ethnocentric, yet a close examination of their linguistic and cultural history reveals a very different picture: although theirs is essentially a monolingual speech community they emerge as a people who have been significantly influenced by other languages and cultures for at least 2000 years. In this primarily sociolinguistic study Professor Loveday takes an eclectic approach, drawing on insights from other subfields of linguistics such as comparative and historical linguistics and stylistics, and from a number of other disciplines - cultural anthropology, social psychology and semiotics. Focusing in particular on the influence of Chinese and English on Japanese, and on how elements from these languages are modified when they are incorporated into Japanese, Professor Loveday offers a general model for understanding language contact behaviour across time and space. The study will be of value to those in search of cross-cultural universals in language contact behaviour, as well as to those with a particular interest in the Japanese case.

The State of theArte: Ecclesiastical Literature on the Northern Frontier of New Spain

An examination of the grammars of indigenous languages of northern Mexico compiled by Jesuit and Franciscan missionaries during the colonial period reveals that the Latinate grammar model used was not slavishly followed, as is often assumed. Moreover, these grammar and related materials provide the groundwork for contributions to disciplines other than linguistics' such as the study of culture change, historiography, and ethnohistory.The heuristic use of Latinate grammatical theory in the grammars sampled here reflects the pragmatic concerns of the missionary confronted with a monolingual speech community. Grammatical theory was as essential frontier technology as medicine, architecture and civil engineering. A person in an initial contact situation needed an efficient means of learning the new language and recording it for posterity. Latinate grammatical theory was used to expedite communication along the expanding frontier.

Bilinguals in a monolingual and a bilingual speech mode: the effect on lexical access.

The time course of lexical access in fluent Portuguese-English bilinguals and in English speaking monolinguals was examined during the on-line processing of spoken sentences using the phoneme-triggered lexical decision task (Blank, 1980). The bilinguals were tested in two distinct speech modes: a monolingual, English or Portuguese, speech mode, and a bilingual, code-switching, speech mode. Although the bilingual’s lexical decision response times to word targets in the monolingual speech modes were identical to those of the monolingual subjects, their response times to code-switched word targets in the bilingual mode were significantly slower. In addition, the bilinguals took longer to detect nonwords in both the monolingual and bilingual modes. These results confirm that bilinguals cannot totally deactivate their other language when in a monolingual speech mode. It is hypothesized that bilinguals search both lexicons when confronted with nonwords, even when in a totally monolingual mode, and that they search the base-language lexicon before the other lexicon when in a bilingual, code-switching, speech mode.