O conhecimento de duas ou mais línguas é algo muito valorizado no mundo do trabalho, assim, o interesse dos pais para que seus filhos aprendam, desde a mais tenra idade, uma língua estrangeira tem crescido. Para atender às necessidades de um mundo cada vez mais multilíngue, as escolas privadas têm incorporado a educação bilíngue na rotina das crianças. O aprendizado do alemão é visto, frequentemente, pelas famílias como um meio de assegurar aos seus filhos sucesso acadêmico, profissional e econômico no futuro. O alemão conquistou seu espaço em muitas escolas pelo Brasil, tanto no Rio Grande do Sul, estado com o maior número de imigrantes do país, como nos grandes centros urbanos. A Língua Alemã hoje é sinônimo de diferencial e a procura por ela se multiplicou nos últimos 15 anos (UPHOFF, 2011, p. 28). O presente trabalho, intitulado Rabe nicht dividieren: análise da interlíngua presente no processo de aprendizagem de Língua Alemã por crianças em uma escola bilíngue teuto-brasileira1 aborda a aprendizagem de uma nova língua em contexto de aula bilíngue. O objetivo da pesquisa foi identificar e caracterizar a vertente de educação bilíngue adotada na instituição em questão, assim como verificar e analisar os processos característicos da interlíngua que se fazem presentes na aprendizagem da Língua Alemã. Os participantes da pesquisa são alunos de uma turma do 1º ano do Ensino Fundamental e o objeto de análise são as produções orais por eles feitas ao reproduzir histórias ou descrever cenas de um livro com imagens panorâmicas. Os dados foram coletados em cinco momentos ao longo de um ano escolar. Identificou-se, a partir da análise dos dados, a ocorrência de interferências, seguida de omissões, simplificações, supergeneralizações, empréstimos e alguns casos de mistura de línguas. / The knowledge of two or more languages is something highly valued in the world of work, so the interest of parents to have their children learn, from an early age, a foreign language has grown. The insertion of children in a bilingual school context in early childhood is a current issue and has attracted the attention of parents and educators. To meet the needs of an increasingly multilingual world, private schools have incorporated bilingual education into the routine of the children. Learning German is seen by families as a means of assuring their children academic, professional and economic success in the future. German has conquered its space in many schools throughout Brazil, both in Rio Grande do Sul, the state with the largest number of German speaking immigrants in the country, and in large urban centers. The German language is seen nowadays as a differential and its demand has multiplied in the last 15 years (UPHOFF, 2011, p. 28). The present work entitled Rabe nicht dividieren: analysis of interlanguage presents in the German language learning process by children in a German-Brazilian bilingual school approaches the learning of a new language in the context of bilingual classroom. The objective of the research was to identify and characterize the bilingual education strand adopted in the institution in question, as well as to verify and analyze the characteristic interlanguage processes that are present in the learning of the German Language. Participants in the survey are children from the 1st year of Elementary School and the object of analysis is their oral productions by retelling stories or describing scenes from a wimmelbook. Data were collected at five different times throughout a school year. From the analysis of the data, we identified the occurrence of interference, followed by omissions, simplifications, supergeneralizations, loans and some cases of language mixing.

Student Number : 0006815V - MA research report - School of Human and Community Development - Faculty of Humanities / In South Africa, the majority of children are bilingual and little research exists on the cognitive processes bilingual children use to spell. This has far-reaching and challenging implications for cognitive models of spelling. Specifically, bilingualism exhibits a pervasive influence on children’s literacy development (Bialystok, 2002). The majority of research on children’s spelling has been conducted internationally with monolingual English children. From international literature, cognitive processing (simultaneous processing and sequential processing) has been identified as an important area for consideration in the spelling acquisition process of English children (Kaufman & Kaufman, 1983b). Simultaneous processing is important for whole word spelling, whilst sequential processing is important for decoding letter sound correspondences. Cross-linguistic research demonstrates a bias towards one or the other spelling strategy may be tied to the depth of a language’s orthography, possibly due to the different demands the language orthography places on how children learn to spell (Frost et al., 1987; Wimmer & Hummer, 1990, 1994; Goswami et al., 1998). The present study examined the relationship between simultaneous and sequential processing and spelling in Grade 3 monolingual English-speaking children and bilingual Afrikaans-English speaking children at one point in time. Thirty bilingual Afrikaans-English children (Afrikaans first language, English second language) and were learning to spell in Afrikaans and in English simultaneously, and thirty monolingual (English first language) learning to spell in English. Simultaneous and sequential processing subtests of the Kaufman Assessment Battery (K-ABC) were administered to the monolingual and to the bilingual children. Monolingual Englishspeaking children received the English word and non-word spelling tests, while the bilingual Afrikaans-English children were asked to spell English and Afrikaans words and non-words (Klein, 1993). The results suggest that lexical (logographic or simultaneous) and non-lexical (alphabetic or sequential) routes are available in English and Afrikaans, but orthography did exert an influence on cognitive processing strategies. Sequential processing demonstrates a higher relationship than simultaneous processing with spelling in English and Afrikaans, although sequential processing contributes more to spelling in a shallow orthography, because the reliable relationship between spelling supports easier and faster computation than in an opaque orthography. Additionally, the results demonstrate that in the bilingual Afrikaans-English children spelling in a second language (L2) rely on spelling skills in a first language (L1), even when the same teaching strategies are used for spelling instruction. Orthography as a tool of academic literacy instruction, influences whether the transfer of spelling skills has a positive or negative influence on spelling in English as a second language in bilingual Afrikaans-English children with a transparent L1. A dual-route model that incorporates the influence of orthographic depth is supported (Seymour, Bunce & Evans, 1992). The present research study concludes that (1) simultaneous processing and sequential processing influence and predict the production of spelling in L1 and L2 in both English and Afrikaans alphabetic orthographies that differ in orthographic transparency, (2) orthographic demands of learning to spell in different orthographies varies and influences cognitive processing resources and decoding skills, which may provide an indication of a cumulative or challenging development of L2 spelling skills particularly when the L1 is transparent. The present research has implications for assessment, traditional spelling models and teaching bilingual children learning to spell in a second language, which is orthographically opaque relative to their transparent mother tongue.

No description available.

Children who speak limited English are at particular jeopardy of school failure because of multiple risk factors. In the later grades, these children have difficult y making progress towards state standards in English reading and English writing. Research with bilingual children indicates that children transfer phonological awareness and writing skills across languages. However, the research on cross-linguistic transfer of early writing is sparse when compared to the phonological awareness research base. This study is important because it aims to address the gap in the literature by exploring ELLs' pathway from Spanish (L1) to English (L2) written language and moderators of this pathway. Using a sample of 110 preschool English-language learners, the children's early writing performance was compared to national norms in 2007 and 2008 using a standardized instrument that prompts them to write letters and words from dictation. The data was analyzed using commonality regression analysis and canonical correlation to examine 1) shared and unique variance of performance on the English dictation measure accounted for by English and Spanish phonological awareness 2) shared and unique variance of performance on the Spanish dictation measure accounted for by English and Spanish phonological awareness and 3) interrelationships between early writing and phonological awareness in English and Spanish. Although it was expected that the student's performance on the English dictation task would be below average when compared to national norms, the students' performance was low average. The results from commonality regression and canonical correlation analysis indicated that the greatest unique contribution to English and Spanish dictation in 2008 was Spanish dictation in 2007. Finally, the results from the canonical correlation regression indicated that the Spanish literacy skills made a greater contribution to the phonological awareness and dictation synthetic variable than did the parallel English literacy skills.

Cross-linguistic transfer has been found to exist, at different degrees, in the process of second/foreign language acquisition. Both the level of orthographic depth and the orthographic distance between L1 and L2 affect the rate of the word learning process in language acquisition. Compared to English, Chinese orthography encodes morphemes within and via characters, lacks inflectional affixation, and contains a significant number of compound words. Extensive morpho-syllabic characters may develop Chinese readers' morphological awareness, as well as their vocabulary. This study examined how Taiwanese children apply decoding skills they have developed in learning Chinese, while they read English words. The degrees of transfer from Chinese to English within a structural model of morphological awareness are examined and analyzed. In addition, moderation effects resulting from regional differences and the length of time spent learning English are also discussed. Students' morphological awareness is measured in six subtests which individually emphasize inflectional morphemes, derivational morphemes, and morphemes of compound words in Chinese and English, respectively. The results show that middle school students' Chinese morphological awareness facilitates their English morphological awareness development. These students also demonstrated the ability to detect morphemes in English, but only to a limited extent. They were skilled at decoding genuine compound words and were able to detect high frequency morphemes within pseudo-words. Their lengthy experience with morphologically complex characters and words in Chinese was most likely the major factor leading to such skill. Finally, in the development of English morphological awareness, the total length of time spent learning English and the length of formal English instruction experienced at the junior high school level were both found to positively correlate with the level of students' English morphological awareness, but not with the length of formal English instruction at the elementary school level. Such a discrepancy can be attributed to the current curriculum prevalent in elementary school which is only aimed at developing students' listening and speaking skills.

This dissertation examines the production, perception and processing of the Catalan-specific mid-vowel categories (/e/-/[open-mid front unrounded vowel]/ and /o/-/[open-mid back rounded vowel]/) by early Spanish-Catalan bilinguals in Majorca (Spain). The analyses focus on the lexical as well as the segmental levels to analyze cognate effects in the production and lexical representations of these early bilinguals, and they explore how their production and perception abilities are related. This study provides evidence that early and highly proficient Spanish-Catalan bilinguals in Majorca maintain two independent phonetic categories in the Catalan mid-vowel space. The first significant finding is that production patterns in Majorca differ from those previously reported in Barcelona, as the Catalan mid-vowel contrasts are not merging into a single Spanish-like mid-vowel for either Catalan-dominants or Spanish-dominants. Additionally, these bilinguals are not 'deaf' to the Catalan-specific mid-vowel contrasts: both language dominance groups perceive the contrast between the Catalan mid-vowel categories despite the overlap with one phonetic category in Spanish. Even though Spanish-dominant bilinguals as a whole are indistinguishable from Catalan-dominant bilinguals in the perception and production tasks, they are found to have a higher error rate in the lexical decision task. The comparison of the acoustic properties of the target vowels in Catalan cognate and non-cognate experimental items reveals that the production of the mid-vowels is affected by cognate status, and that these cognate effects are also found in the word recognition of aurally presented stimuli. Finally, bilinguals who produced the mid-vowels with a smaller Euclidean distance are more likely than bilinguals who maintain a more robust contrast in their productions to have a higher error rate in the AXB discrimination and lexical decision tasks. The present study contributes to the discussion regarding the organization of early bilinguals' dominant and non-dominant phonetic systems, and implications are considered for cross-linguistic models of bilingual speech production and perception. It is proposed that the exemplar model of lexical representation (Bybee, 2001; Pierrehumbert, 2001) can be extended to include bilingual lexical connections that can account for the interactions between the phonetic and lexical levels of early bilingual individuals. / text

This thesis presents the results of an experimental study designed to examine whether children raised bilingually in Japanese and English from birth in Christchurch, New Zealand, exhibit the same morphological case and topic marking knowledge in Japanese as monolingual children in Japan. The participants were 34 children aged between five and eleven years who have been raised in a one-person one-language environment in an English dominant community. The study replicated previous studies on monolingual Japanese children, and involved two widely used paradigms for assessing a child’s grammar: picture selection, and elicited imitation. The responses of the children in this study were different from those reported in studies of monolingual children. In the picture selection tasks, some children in this study interpreted the agent-patient relationship based on the word order cue in the object-initial types of transitive sentences, whereas previous studies have demonstrated that monolingual children five years and older are able to interpret the agent-patient relationship in the same way as adults, using the case marking cue. Moreover, in the elicited imitation tasks, many children in this study re-analysed the topic-comment construction as a genitive possessive when the particles in the stimuli were masked with noise. This pattern has not been reported in any previous study. The results also revealed that there was a great degree of individual variation. The study suggests cross-linguistic influence from English on Japanese as a possible explanation for the difference between the children in this study and monolinguals. The phenomena observed in the results satisfies two conditions for cross-linguistic influence proposed by Hulk and Müller (2000) and Müller and Hulk (2001), because (i) English and Japanese overlap at the surface level in terms of the agent position in a canonical sentence and the possessive structure, and (ii) the problematic structures for some children in this study involved the interface between syntax and pragmatics in the C-domain. The study, however, has no principled explanation for the individual variation found because of a lack of data on the Japanese input and the child’s fluency, both of which are likely to affect simultaneous bilingual development.

This study was an investigation into the aural and written receptive knowledge of the English semantics of English lexis that is loanwords in the Japanese language and the predictive strength of the variables of semantic distance, concreteness/abstractness, polysemy, phonological distance, number of syllables, number of phonemes, number of letters, part of speech (POS), English Frequency, and frequency in Japanese in relation to accurate semantic knowledge. The participants (N = 215) were first- and second-year, non-English majors at a large university in Western Japan. The participants were from 10 intact English classes focused on reading, writing, and communication skills. Data were collected using eight instruments: the Listening Vocabulary Levels Test, Aural Loanword Test, Aural Non-Loanword Test, New Vocabulary Levels Test, Written Loanword Test, Written Non-Loanword Test, and Japanese Loanword Frequency Rating Task. Additionally, data were collected from five Japanese L1 speakers highly proficient at English on the Semantic Distance Rating Task. The data were first analyzed using the Rasch dichotomous model to examine instrument reliability and validity as well as to transform the data into Rasch person ability estimates and Rasch item ability estimates. Pearson correlations were used to determine the strength of the relationship between loanwords and non-loanwords. Repeated-measures ANOVA—with follow up t-tests were used to determine the differences between the four semantic tests: the Aural Loanword Test, the Aural Non-Loanword Test, the Written Loanword Test, and the Written Non-Loanword Test. Four multiple linear regression analyses were conducted using the predictor variables semantic distance, concreteness/abstractness, phonological distance, number of syllables, number of phonemes, number of letters, part of speech, English Frequency, and frequency in Japanese. The results of the Pearson analyses showed strong correlations between the aural and written loanword and non-loanword measures. This finding indicated that the participants’ knowledge of loanwords was relatively equivalent to their knowledge of non-loanwords. The results of the comparison between aural and written loanword knowledge showed that written knowledge of loanwords was greater than aural knowledge of loanwords. Further comparisons between the loanword and non-loanword tests showed that receptive aural non-loanword knowledge was greater than aural loanword knowledge, and that written non-loanword knowledge was greater than written loanword knowledge. These comparisons showed that English semantic knowledge of loanwords was less accurate than that of non-loanwords, which implied that the accurate acquisition of English semantic knowledge of loanwords was impeded by Japanese L1 lexical knowledge. The results of the multiple regressions indicated that the only substantial predictor of lexical acquisition for both loanwords and non-loanwords in both modalities (aural and written) was English Frequency. Although the effect size of English frequency was substantial, it was less so on the aural and written loanword measures. This finding implied that English linguistic gains of repeated exposure were most likely muted by entrenched L1 semantic knowledge. Overall, the results showed that loanwords are generally acquired with greater difficulty than non-loanwords and that they should not always be considered a form of receptive knowledge of English lexis. / Applied Linguistics

Thesis (PhD)--Stellenbosch University, 2013. / ENGLISH ABSTRACT: The dissertation is a comparative analysis of the structural (generic/cognitive) and ideological properties of Zimbabwean news reports in English, Shona and Ndebele, focusing specifically on the examination of the proliferation of authorial attitudinal subjectivities in ‘controversial’ ‘hard news’ reports and the ‘objectivity’ ideal. The study, thus, compares the textuality of Zimbabwean printed news reports from the English newspapers (The Herald, Zimbabwe Independent and Newsday), the Shona newspaper (Kwayedza) and the Ndebele newspaper (Umthunywa) during the period from January 2010 to August 2012. The period represents an interesting epoch in the country’s political landscape. It is a period characterized by a power-sharing government, a political situation that has highly polarized the media and as such, media stances in relation to either of the two major parties to the unity government, the Zimbabwe African National Union Patriotic Front (ZANU PF) and the Movement for Democratic Change (MDC-T). Couched in the theoretical explications of Appraisal Theory, specifically the ‘reporter voice’ configuration, the study sought to investigate the proliferation of journalistic ideological subjectivities in ‘hard news’ reports – a genre of news reporting that is largely characterised by claims of ‘objectivity’ and/or ‘neutrality’ and dispassionate journalistic reporting positions. The study, also assuming the orbital structure model developed by Iedema, Feez and White (1994) and White (1997, 1998) in the analysis of ‘hard news’ report in English broadsheet reporting, furthermore sought to investigate whether the textuality and cognitive/rhetorical structure of ‘hard news’ reports in news reports from the three Zimbabwean language journalistic cultures are organised around the same structure. The corpus of news reports analysed in this study were examined for the proliferation of instances of observable authorial ideological positionings by focusing how the choices made in terms of lexical, lexicogrammatical and syntagmatic resources signal evaluative keys that betray authorial ideological subjectivities. The texts were, thus, subjected to close textual analyses in terms of generic structure and journalistic voices. The study shows that Zimbabwean news reports in English, Shona and Ndebele generally share the same structure as expressed by the orbital model, in which authorial subjective evaluations are curtailed through a variety of strategic impersonalisations – largely ‘attribution’. However, despite these similarities, significant differences were observed with regards to the textuality of news reports as well as the uses made of attributed materials. / AFRIKAANSE OPSOMMING: Die verhandeling behels ʼn vergelykende analise van die strukturele (generiese/kognitiewe) en ideologiese eienskappe van Zimbabwiese nuusberigte in Engels, Shona en Ndebele, wat veral op die ondersoek van die proliferasie van subjektiwiteite in die houdings van outeurs by ‘kontroversiële’ ‘hardenuusberigte’ en die ideaal van ‘objektiwiteit’ fokus. Die studie het dus die tekstualiteit van Zimbabwiese gedrukte nuusberigte uit die Engels koerante The Herald, Zimbabwe Independent en Newsday, die Shona-koerant Kwayedza en die Ndebele-koerant Umthunywa uit die tydperk Januarie 2010 tot Augustus 2012 vergelyk. Dié tydperk verteenwoordig ʼn interessante tydvak in die land se politieke landskap. Dit is ʼn tydperk gekenmerk deur ʼn magsdelende regering, ʼn politieke situasie wat die media tot ʼn groot mate gepolariseer het en as sodanig mediastandpunte in verband met enige van die twee belangrikste partye in die eenheidsregering, die Zimbabwe Africa National Union Patriotic Front (ZANU PF) en die Movement for Democratic Change (MDC-T). Uitgedruk in die teoretiese uiteensettings van teorie van waardebepaling, in die besonder die ‘stem van die verslaggewer’-konfigurasie, het die studie gepoog om die uitbreiding van joernalistieke ideologiese subjektiwiteite in ‘hardenuusberigte’ – ʼn beriggewingsgenre wat grootliks deur aansprake van ‘objektiwiteit’ en/of ‘neutraliteit’ en posisies van emosielose joernalistieke beriggewing gekenmerk word – te ondersoek. Die studie, wat ook die orbitale struktuur-model ontwikkel deur Iedema, Feez en White (1994) en White (1997, 1998) by die analise van ‘hardenuusberigte’ in Engelstalige breëbladberiggewing gebruik het, het verder daarna gestreef om ondersoek in te stel daarna of die tekstualiteit en kognitiewe/retoriese struktuur van ‘hardenuusberigte’ in drie joernalistieke kulture in Zimbabwe om dieselfde struktuur heen georganiseer is. Die korpus nuusberigte wat in hierdie studie ontleed is, is nagegaan vir die proliferasie van gevalle van waarneembare ideologiese posisionerings van die skrywers deur te fokus op hoe die keuses wat gemaak is ten opsigte van leksikale, leksiko-grammatikale en sintagmatiese hulpbronne bewys lewer van waardebepalende sleutels wat ideologiese subjektiwiteite van die outeurs verklap. Die tekste was dus onderworpe aan noukeurige tekstuele analises ten opsigte van generiese struktuur en joernalistieke stemme. Die studie het aangetoon dat Zimbabwiese nuusberigte in Engels, Shona en Ndebele in die reël dieselfde struktuur deel as wat deur die orbitale model uitgedruk word, waarin subjektiewe evaluerings deur die outeur beperk word deur ʼn verskeidenheid strategiese onpersoonlikhede – hoofsaaklik ‘toeskrywing’. Ondanks hierdie ooreenkomste is beduidende verskille waargeneem met betrekking tot die tekstualiteit van nuusberigte asook die gebruik wat van toegeskryfde materiaal gemaak word. / Deep gratitude goes to the Graduate School (Faculty of Arts and Social Sciences – University of Stellenbosch) for the funding/scholarship extended to me through the African Doctoral Academy (ADA), which has made this work see the light of day

<p>U ovom istraživanju cilj je da na osnovu lične percepcije fenomena individualne vi&scaron;ejezičnosti kod studenata i studentkinja Filozofskog fakulteta, Univerziteta u Novom Sadu, ispitam njihove stavove prema vlastitoj dvo- i vi&scaron;ejezičnosti i utvrdim da li se, u kojoj meri i na koji način odvijaju međujezički uticaji u procesu učenja trećeg jezika.<br />U Vojvodini je u službenoj upotrebi srpski jezik i ćirilično pismo, mađarski, slovački, hrvatski, rumunski i rusinski jezik i njihova pisma (Statut AP Vojvodine (&bdquo;Službeni list APV&rdquo; br.17/09)) u skladu sa zakonom i pokrajinskom skup&scaron;tinskom odlukom, a na Filozofskom fakultetu u Novom Sadu postoji duga tradicija istraživanja dvo- i vi&scaron;ejezičnosti u odnosu na jezik većinskog naroda (Mike&scaron; 1961, 1964, 1965, 1991; 1998; Mike&scaron; i Vlahović 1966; 1967; Mike&scaron; i Savić 1972, 1973, 1974; Genc 1973, 1979, 1981, 1985, 1991; Međe&scaron;i 2009; Tir 2002; Horakova 2002; Puja-Badesku 2009).<br />Korpus empirijskih podataka čine dve grupe odgovora: odgovori dvojezičnih studenata (26) i studentkinja (74) prve godine Filozofskog fakulteta u Novom Sadu &scaron;kolske 2011/12. godine (prosečna starost 21;8 godina) za koje su dobijeni podaci pomoću upitnika i odgovori dve trojezične alumnistkinje istog fakulteta (44 godine) za koje su dobijeni podaci audio zapisom njihovih ličnih ispovesti (životnih priča).<br />Upitnik ima sledeće celine: o maternjem jeziku, drugom i trećem jeziku, kao i međujezičkim uticajima; a audio zapisom su zabeleženi odgovori na slična pitanja u polustrukturiranom intervjuu. Prikupljeni podaci su analizirani na osnovu kombinacije mađarskog kao ugro-finskog i slovenskih (srpski/hrvatski, rusinski, slovački, ukrajinski, poljski, ruski), germanskih (nemački, engleski) i romanskih (rumunski, francuski, italijanski, &scaron;panski) kao idoevropskih jezika.<br />Dobijeni rezultati potvrđuju rezultate savremenih istraživanja dvo- i vi&scaron;ejezičnosti i pokazuju sledeće: bez obzira na tipolo&scaron;ke karakteristike jezika dvojezičnost efikasno utiče na usvajanje trećeg jezika (Ringbom 1987; Cenoz 2001; Jessner 2007; Gibbons 2009), &scaron;to je u ovom istraživanju dokumentovano za kombinacije jezika: mađarski kao ugro-finski i slovenski (srpski/hrvatski, rusinski, slovački, ukrajinski, poljski, ruski), germanski (nemački, engleski) i romanski (rumunski, francuski, italijanski, &scaron;panski) kao idoevropski jezici; drugi jezik ima veći uticaj na učenje trećeg jezika od prvonaučenog ili usvojenog jezika (Williams and Hammarberg 1998; Bardel and Falk 2007), ukoliko je kompetencija na visokom nivou, a izloženost drugom jeziku česta i dugotrajna i ukoliko su drugi i treći jezik tipolo&scaron;ki slični; transfer jezičkog znanja sprovodi se na nesvesnom nivou i to če&scaron;će u neformalnim, nego u formalnim situacijama, kao &scaron;to je potvrđeno i u drugim istraživanjima (Hoffmann 2007); komunikativna strategija opisivanja se če&scaron;će primenjuje u procesu stvaranja iskaza na trećem jeziku od strategije pravljenja hibridnih varijanti; kvalitativna analiza odgovora iz životnih priča potvrđuje osnovne rezultate dobijene na osnovu odgovora studenata i studentkinja i otkriva niz nijansi koje nije moguće dobiti upitnikom, a koje se odnose na kontekst, lična iskustva i sposobnosti da se metajezičke sposobnosti stave u funkciju učenja jezika.<br />Iako je trojezičnost po svemu sudeći budućnost civilizacijskog toka i obrazovnog sistema, možda čak i od pred&scaron;kolskog uzrasta, o ovom fenomenu ima jo&scaron; uvek malo literature. Na bazi rezultata dobijenih u ovom istraživanju, iako njihova pouzdanost nije apsolutna, jer zavise od ličnog procenjivanja fenomena, moguće je nanovo razmotriti postojeće teorije i ponovo opisati metalingvističke sposobnosti kada su u pitanju jezici koji nisu srodni u odnosu na jezike koji su srodni kada je u pitanju trojezičnost u Vojvodini.</p> / <p>The aim of this study is to investigate<br />attitudes towards personal bilingualism and<br />multilingualism and to determine if there is a<br />cross-linguistic influence in the third<br />language acquisition process, how and how<br />often it occurs, according to a personal<br />perception of the phenomenon of the<br />individual multilingualism among students at<br />the Faculty of Philosophy, University of Novi<br />Sad.<br />In Vojvodina, the officially used languages<br />are: Serbian, Hungarian, Slovak, Croatian,<br />Romanian and Ruthenian (AP of Vojvodina<br />Statute (&bdquo;Službeni list APV&rdquo; br.17/09)). At<br />the Faculty of Philosophy in Novi Sad<br />bilingualism, multilingualism and the<br />majority language has been traditionally<br />studied since its foundation (Mike&scaron; 1961,<br />1964, 1965, 1991; 1998; Mike&scaron; i Vlahović<br />1966; 1967; Mike&scaron; i Savić 1972, 1973, 1974;<br />Genc 1973, 1979, 1981, 1985, 1991; Međe&scaron;i<br />2009; Tir 2002; Horakova 2002; Puja-<br />Badesku 2009).<br />The empirical data include two groups of<br />answers: answers that the first year (2011/12)<br />bilingual female (74) and male (26) students<br />(mean age 21;8) gave to questions about their<br />mother tongue, second and third language and<br />cross-linguistic influence in a questionnaire<br />and answers that the trilingual alumni of the same faculty (44 years old) gave in their oral</p><p>history.<br />The questionnaire includes questions about<br />mother tongue, second and third language, as<br />well as about cross-linguistic influences. The<br />questions in the semi-structured interview for<br />the oral history are about the same topics. The<br />obtained data are analyzed according to the<br />combination of Hungarian as Finno-Ugric<br />language and Slavic (Serbian/Croatian,<br />Ruthenian, Slovak, Ukrainian, Polish,<br />Russian), Germanic (German, English) and<br />Romance languages (Romanian, French,<br />Italian, Spanish) as Indo-European languages.<br />The obtained results confirm the<br />contemporary findings in bilingualism and<br />multilingualism research and show the<br />following: bilingualism has a positive<br />influence on third language acquisition<br />regardless of their typological characteristics<br />(Cenoz 2001; Jessner 2007) and in this<br />research it is confirmed for combination of<br />languages: Hungarian as Finno-Ugric<br />language and Slavic (Serbian/Croatian,<br />Ruthenian, Slovak, Ukrainian, Polish,<br />Russian), Germanic (German, English) and<br />Romance languages (Romanian, French,<br />Italian, Spanish) as Indo-European languages;<br />the second language has a greater influence<br />on the third language acquisition process than the first one (Williams and Hammarberg</p><p>1998; Bardel and Falk 2007), if the<br />competence is on the high level, the second<br />language exposure frequent and long, and if<br />the second and the third language are<br />typologically similar; the linguistic<br />knowledge is transferred subconsciously,<br />more often in informal than in formal<br />situations, as it is confirmed in other studies<br />(Hoffmann 2007); description as a<br />communicational strategy is more often used<br />than creation of hybrid forms; qualitative<br />analysis of the obtained answers in oral<br />histories confirm results obtained by the<br />questionnaire, but it also reveals various<br />nuances that cannot be found using<br />questionnaire, about the context, personal<br />experiences and abilities to use metalinguistic<br />awareness in the language learning process.<br />In spite of the fact that trilingualism is the<br />future in the civilization evolution and in the<br />educational system, even starting at the<br />preschool level, there is not enough literature<br />about the phenomenon. According to the<br />results obtained in this research, although<br />they are not absolutely reliable, since they<br />depend on a personal judgment of the<br />phenomenon, it is possible to redefine<br />existing theories and to describe<br />metalinguistic abilities again when</p>