The permeability of tag questions in a language contact situation: The case of Spanish-Portuguese bilinguals

Ana M. Carvalho and Joseph Kern

Abstract

In this paper, we utilize negative polarity tag questions in order to assess to what extent discourse-pragmatic variables are susceptible to language contact induced changes. Based on a comparison of forms and functions of negative tags in the varieties spoken by Portuguese-Spanish bilinguals in a community on the Uruguayan-Brazilian border with the one spoken by monolinguals in the Uruguayan capital, we aimed at assessing to what extent any differences in this variable behavior may be affected by contact with Portuguese. Our results indicate that, despite the high permeability of discourse-pragmatic features in contact situations attested in the literature and the presumed tendency for cognate languages to converge, the forms and functions of negative tags in bilingual Spanish did not radically differ from the monolingual variety. We found, instead, an intricate pattern of convergences and divergences that challenges the presupposed assumptions about extreme permeability of cognate discourse pragmatic systems in contact.

Keywords:
Publication history
Table of contents

1.Introduction

There is plenty of evidence by now that discourse-pragmatic features can be extremely permeable and susceptible to borrowing in bilingual discourse (Dajko and Carmichael 2014Dajko, Nathalie, and Katie Carmichael 2014 “But Qui C’est La Différence ? Discourse Markers in Louisiana French: The Case of but vs. mais.” Language in Society 43 (2): 159–83. CrossrefGoogle Scholar; Myers-Scotton and Jake 1995Myers-Scotton, Carol, and Janice L. Jake 1995 “Matching Lemmas in a Bilingual Language Competence and Production Model: Evidence from Intrasentential Code Switching.” Linguistics 33 (5): 981–1024. CrossrefGoogle Scholar, among others). This counters the generalization that only content words such as nouns, verbs, and adjectives are likely to be borrowed in contact situations. By framing discourse organization as opposed to participating in sentence formation, discourse-pragmatic features may be expected to be highly permeable since they can be freely incorporated in the discourse structure of other languages by bracketing units of talk. This permeability is hypothesized to be even more robust in situations in which cognate languages are in prolonged contact, since it is expected that congruent structures and lexical similarities facilitate borrowing behavior (Muysken 2000Muysken, Pieter, Carmen Pena Díaz, and Pieter Cornelis Muysken 2000Bilingual Speech: A typology of Code-mixing. Vol. 11. Cambridge, UK: Cambridge University Press.Google Scholar). In this study, we aim at testing this assumption by analyzing the use of a specific discourse-pragmatic feature, negative polarity tag questions, among bilinguals in two cognate languages, Spanish and Portuguese. In doing so, we compare the use of negative tags in Spanish among bilinguals from Rivera, a Uruguayan town near the Brazilian border, with their use among monolingual Spanish speakers from the Uruguayan capital, Montevideo. In addition, we compare the use of negative tags in the Spanish and Portuguese varieties used by the same bilinguals. Our main goal is to assess the extent to which the negative tag system in bilingual Spanish diverges from its monolingual counterpart due to contact with Portuguese.

Examples that illustrate the “pragmatic detachability” (Matras 1998Matras, Yaron 1998 “Utterance Modifiers and Universals of Grammatical Borrowing.” Linguistics 36 (2): 281–332. CrossrefGoogle Scholar) of discourse-pragmatic features in language contact situations are abundant, as seen in the incorporation of English discourse markers in U.S. Spanish (Aaron 2004Aaron, Jessi 2004 ““So Respetamos Un Tradicion Del Uno Al Otro.” So and Entonces in New Mexican Bilingual Discourse.” Spanish in Context 1 (2): 161–79. CrossrefGoogle Scholar; Torres 2002Torres, Lourdes 2002 “Bilingual Discourse Markers in Puerto Rican Spanish.” Language in Society 31 (1): 65–83. CrossrefGoogle Scholar; Torres and Potowksi 2008Torres, Lourdes, and Kim Potowski 2008 “A Comparative Study of Bilingual Discourse Markers in Chicago Mexican, Puerto Rican, and MexiRican Spanish.” International Journal of Bilingualism 12 (4): 263–279. CrossrefGoogle Scholar), in U.S. French (Dajko and Carmichael 2014Dajko, Nathalie, and Katie Carmichael 2014 “But Qui C’est La Différence ? Discourse Markers in Louisiana French: The Case of but vs. mais.” Language in Society 43 (2): 159–83. CrossrefGoogle Scholar), in U.S. German (Fuller 2001Fuller, Janet M. 2001 “The Principle of Pragmatic Detachability in Borrowing: English-origin Discourse Markers in Pennsylvania German.” Linguistics 39 (2): 351–370. ISSU 372. CrossrefGoogle Scholar; Salmons 1990Salmons, Joe 1990 “Bilingual Discourse Marking: Code Switching, Borrowing, and Convergence in Some German-American Dialects.” Linguistics 28 (3): 453–480. CrossrefGoogle Scholar), and in Canadian French (Mougeon and Beniak 1991Mougeon, Raymond, and Edouard Beniak 1991Linguistic Consequences of Language Contact and Restriction: The Case of French in Ontario, Canada. Oxford: Clarendon.Google Scholar; Sankoff et al. 1997Sankoff, Gillian, Pierrette Thibault, Naomi Nagy, Hélène Blondeau, Marie-Odile Fonollosa, and Lucie Gagnon 1997 “Variation in the Use of Discourse Markers in a Language Contact Situation.” Language Variation and Change 9 (2): 191–217. CrossrefGoogle Scholar). The high permeability of discourse-pragmatic features in language contact situations may lead to the incorporation of borrowed discourse markers that result in the displacement of native ones (Dajko and Carmichael 2014Dajko, Nathalie, and Katie Carmichael 2014 “But Qui C’est La Différence ? Discourse Markers in Louisiana French: The Case of but vs. mais.” Language in Society 43 (2): 159–83. CrossrefGoogle Scholar; Salmons 1990Salmons, Joe 1990 “Bilingual Discourse Marking: Code Switching, Borrowing, and Convergence in Some German-American Dialects.” Linguistics 28 (3): 453–480. CrossrefGoogle Scholar), or they may be added to the recipient language’s repertoire in order to fulfill discourse functions not fulfilled by native discourse markers, as seen by Hlavac (2006)Hlavac, Jim 2006 “Bilingual Discourse Markers: Evidence from Croatian-English Code-switching.” Journal of Pragmatics 38 (11): 1870–1900. CrossrefGoogle Scholar in his study of English-Croatian bilinguals in Australia. In fact, discourse-pragmatic features are so highly permeable that they may even enter monolingual speech, as seen in the English-originated pliis in Finnish (Peterson and Vaattovaara 2014Peterson, Elizabeth, and Johanna Vaattovaara 2014 “Kiitos and pliis: The Relationship of Native and Borrowed Politeness Markers in Finnish.” Journal of Politeness Research 10 (2): 247–269. CrossrefGoogle Scholar), Spanish bueno in Gaucho Portuguese in Southern Brazil (Leite and Martins 2006Leite, Ligia Chiappini Moraes, and Maria Helena Martins 2006Cone Sul: Fluxos, Representações e Percepções Vol. 38. Editora Hucitec.Google Scholar), and the well documented widespread use of English ‘OK’ in so many languages.

“Pragmatic borrowing” (Andersen 2014 2014 “Pragmatic Borrowing.” Journal of Pragmatics 67: 17–33. CrossrefGoogle Scholar) in situations of prolonged language contact may lead to the diachronic convergence of systems of discourse-pragmatic features, as seen in U.S. German by Salmons (1990)Salmons, Joe 1990 “Bilingual Discourse Marking: Code Switching, Borrowing, and Convergence in Some German-American Dialects.” Linguistics 28 (3): 453–480. CrossrefGoogle Scholar. From a synchronic viewpoint, Matras (2000) 2000 “Fusion and the Cognitive Basis for Bilingual Discourse Markers.” International Journal of Bilingualism 4 (4): 505–528. CrossrefGoogle Scholar claims that systems may completely “fuse” and result in a “wholesale, class-specific nonseparation of the systems” (Matras 2000 2000 “Fusion and the Cognitive Basis for Bilingual Discourse Markers.” International Journal of Bilingualism 4 (4): 505–528. CrossrefGoogle Scholar, 512). Such a fusion is triggered by bilinguals’ need to reduce the cognitive load by eliminating language-specific options resulting in unintentional bilingual “slips” (Matras 2000 2000 “Fusion and the Cognitive Basis for Bilingual Discourse Markers.” International Journal of Bilingualism 4 (4): 505–528. CrossrefGoogle Scholar, 519). Interestingly, while it is well known that content words are usually borrowed from the majority language and incorporated into the minority language, the pragmatic detachability of discourse-pragmatic features allows for frequent incorporation of minority language particles into the recipient majority language as well. This pattern is illustrated by several examples presented by Lim and Ansaldo’s (2016)Lim, Lisa, and Umberto Ansaldo 2016Languages in Contact. Cambridge, UK: Cambridge University Press. CrossrefGoogle Scholar careful examination of the incorporation of particles from several Sinitic substrates (Mandarim, Cantonese, Bazzar Nakat, and Hokkien) into Asian English varieties, leading them to conclude that discourse particles are widespread features of contact-induced change in contact varieties in general (Lim and Anasaldo 2016Lim, Lisa, and Umberto Ansaldo 2016Languages in Contact. Cambridge, UK: Cambridge University Press. CrossrefGoogle Scholar, 110).

The analysis of discourse-pragmatic features to examine language contact-induced changes in the speech of Spanish-Portuguese bilinguals on the Uruguayan-Brazilian border can contribute to the scholarship on this language contact situation. While some scholars claim that structural similarities between Spanish and Portuguese can lead to congruent lexicalization and the creation of a new, Spanish/Portuguese mixed code in these border communities that renders the separation of languages impossible (Lipski 2006Lipski, John M. 2006 “Too Close for Comfort? The Genesis of “Portuñol/Portunhol”.” In Selected proceedings of the 8th Hispanic Linguistics Symposium 1: 22–54. Somerville: Cascadilla Proceedings Project.Google Scholar, 2009 2009 “Searching for the Origins of Uruguayan Fronterizo Dialects: Radical Code-mixing as “Fluent Dysfluency”. Journal of Portuguese Linguistics 8 (1): 3–44. CrossrefGoogle Scholar; Marín 2001Marín, Francisco Marco 2001 “De lenguas y fronteras. El Spanglish y el portuñol.” Nueva Revista de Política, Cultura y Arte, 74: 72–79.Google Scholar; Sturza 2004Sturza, Elaine 2004 “Fronteiras e Práticas Linguísticas: Um Olhar sobre o Portunhol.” Revista Internacional de Linguística Iberoamericana 3: 151–162.Google Scholar), others have argued from a variationist perspective that both Spanish and Portuguese coexist in a situation of diglossic bilingualism. Although lexical borrowings are frequent in both languages, their phonological and morphsyntactic structures remain separate. For example, multiple analyses of Uruguayan Portuguese spoken by bilinguals point to strong continuities between this dialect and Brazilian Portuguese as opposed to convergence towards Spanish (Carvalho 2003Carvalho, Ana Maria 2003 “The Sociolinguistic Distribution of (lh) in Uruguayan Portuguese: A Case of Dialectal Diffusion.” In Linguistic Theory and Language Development in Hispanic languages: Papers from the 5th Hispanic Linguistics Symposium and the 4th Conference on the Acquisition of Spanish and Portuguese, ed. by Silvina Montrul & Franscisco Ordóñez, 30–43. Somerville, MA: Cascadilla Press.Google Scholar, 2004 2004 “I Speak like the Guys on TV: Palatalization and the Urbanization of Uruguayan Portuguese.” Language Variation and Change 16 (2): 127–151. CrossrefGoogle Scholar, 2016 2016 “The Analysis of Languages in Contact: A Case Study through a Variationist Lens. Cadernos de Estudos Linguísticos 58 (3): 401–424. CrossrefGoogle Scholar; Carvalho and Bessett 2015Carvalho, Ana Maria and Bessett, Ryan 2015 “Subject Pronoun Expression among Spanish-Portuguese Bilinguals.” In Subject Personal Pronouns in Spanish: a Cross-dialectal Perspective, ed. by Ana Maria Carvalho, Rafael Orozco and Naomi Lapidus Shin, 275–315. Washington D.C.: Georgetown University Press.Google Scholar; Castañeda-Molla 2011Castañeda-Molla, Rosa Maria. Linguistic Variation in a Border Town: Palatalization of Dental Stops and Vowel Nasalization in Rivera. PhD diss., University of Florida 2011.Google Scholar; Córdoba 2017Córdoba, Alexander Severo 2017 “O Português Uruguai Falado em Tranqueras-Uruguai: O Fenômeno da Elevação das Vogais Postônicas Finais.” Web Revista SOCIODIALETO 7 (20): 389–427.Google Scholar; Meirelles 2009Meirelles, Virginia Garrido 2009 “O Português da Fronteira Uruguai-Brasil.” In Português em Contato, ed. by Ana M. Carvalho, 257–76. Madrid: Iberoamericana. CrossrefGoogle Scholar; Pacheco 2017Pacheco, Cintia da Silva 2017 “Identidade sociolinguística na fronteira de Aceguá (Brasil-Uruguai).” Revista de Estudos da Linguagem 25 (1): 276–304. CrossrefGoogle Scholar), presenting evidence against the idea that systems will converge in situations of prolonged contact between cognate languages. The Spanish variety spoken by bilinguals on the border has been studied as well from a variationist view point by Waltermire (2006)Waltermire, Mark 2006Social and Linguistic Correlates of Spanish-Portuguese Bilingualism on the Uruguayan-Brazilian border. New Mexico: ProQuest.Google Scholar and Carvalho (2006 2006 “Spanish (s) Aspiration as a Prestige Marker on the Uruguayan-Brazilian Border.” Spanish in Context 3 (1): 85–114. CrossrefGoogle Scholar, 2010 2010 “¿Eres de la frontera o sos de la capital? Variation and Alternation of Second-person Verbal Forms in Uruguayan Border Spanish.” Southwest Journal of Linguistics 29 (1): 1–24.Google Scholar) who show no evidence that border Spanish significantly diverges from the variety spoken by monolinguals in the Uruguayan capital of Montevideo. By showing clear similarities between border varieties and their monolingual counterparts, these studies also show a lack of convergence due to the independence of linguistic systems despite prolonged contact.

To our knowledge, discourse-pragmatic features have not been studied in these bilingual communities, thus presenting a test bed for the exploration of the effects of language contact on these pragmatic systems. In this paper, we utilize one type of discourse-pragmatic variable, namely negative polarity tag questions, in order to assess to what extent discourse-pragmatic variables are susceptible to language contact induced-changes among Spanish-Portuguese bilinguals in Uruguay. Based on a detailed comparison of the forms and functions of negative tags in the Spanish variety spoken by Portuguese-Spanish bilinguals with the one spoken by Spanish monolinguals, we aim at assessing to what extent any differences in this variable behavior may be affected by contact with Portuguese. To complement our assessment, we include an analysis of the forms and functions of negative tags in the Portuguese variety spoken by the same Spanish-Portuguese bilinguals. Our findings demonstrate that despite the high permeability of discourse-pragmatic features in contact situations attested in the literature, the pre-contact similarities between Spanish and Portuguese, and the presumed tendency for cognate languages to converge, the forms of negative polarity tag questions in bilingual Spanish did not radically differ from the forms in the monolingual variety spoken in the Uruguayan capital, with a few exceptions, indicating an intricate pattern of convergences and divergences that challenge the presupposed assumptions about extreme permeability of cognate discourse-pragmatic systems in contact.

2.The community

The long-term coexistence of Spanish and Portuguese in bilingual communities in northern Uruguay presents an opportunity to assess the effects of contact between cognate languages’ repertoires of discourse-pragmatic features. Spanish and Portuguese have coexisted in a somewhat diglossic situation in several communities in northern Uruguay, along the border with Brazil. Historically, Portuguese was the language of settlers in the region until the end of the nineteenth century, when national borders were demarcated and the Uruguayan government implemented several nation-building initiatives in the Uruguayan territory. Spanish-only language policies were then implemented through the opening of several Spanish-speaking schools and public offices along the border, and Spanish-Portuguese bilingualism gradually became a reality that prevails to today in border towns such as Rivera, Artiga, and Aceguá and their surrounding communities (Elizaincín 1995Elizaincín, Adolfo 1995 “Personal Pronouns for Inanimate Entities in Uruguayan Spanish in Contact with Portuguese.” In Spanish in Four Continents: Studies in Language Contact and Bilingualism, ed. by Carmen Silva-Corvalán, 117–131. Washington, D.C.: Georgetown University Press.Google Scholar). In these communities, Spanish is the prestigious language preferred in public domains, while Portuguese is the minority language usually used in in-group situations. Studies that have investigated the social stratification of language choice have indicated that while Spanish is prevalent in middle-class households, Portuguese is favored by lower socio-economic groups. In general, these communities are considered cases where societal Spanish-Portuguese bilingualism is widespread, where both languages are usually used on a daily basis (Carvalho 2003Carvalho, Ana Maria 2003 “The Sociolinguistic Distribution of (lh) in Uruguayan Portuguese: A Case of Dialectal Diffusion.” In Linguistic Theory and Language Development in Hispanic languages: Papers from the 5th Hispanic Linguistics Symposium and the 4th Conference on the Acquisition of Spanish and Portuguese, ed. by Silvina Montrul & Franscisco Ordóñez, 30–43. Somerville, MA: Cascadilla Press.Google Scholar, 2016 2016 “The Analysis of Languages in Contact: A Case Study through a Variationist Lens. Cadernos de Estudos Linguísticos 58 (3): 401–424. CrossrefGoogle Scholar; Behares 1984Behares, Luiz Ernesto 1984 “Diglosia en la Sociedad Escolar de la Frontera Uruguaya con Brasil: Matriz Social del Bilinguismo.” Cadernos de Estudos Linguísticos 6: 229–235.Google Scholar; Elizaincín et al. 1987Elizaincín, Adolfo, Luis Ernesto Behares, and Graciela Barrios 1987Nos falemo brasilero: dialectos portugueses en Uruguay. Montevideo, Uruguay: Editorial Amesur.Google Scholar; Waltermire 2011 2011 “Frequency Effects on the Morphological Conditioning of Syllable-Final /s/ Reduction in Border Uruguayan Spanish.” Journal of Language Contact 4 (1): 26–55. CrossrefGoogle Scholar).

The bilingual data for this study was collected by Carvalho in Rivera, a border town adjacent to its twin city of Santana do Livramento, Brazil (Figure 1). The border is open and pedestrians and drivers may cross the avenue that separates the countries freely and undocumented. A town of approximately 80,000 inhabitants, Rivera is the capital of a province whose main economic activity is cattle. With daily buses to the Uruguayan capital of Montevideo, located 500 kilometers south, several border dwellers maintain contact with monolingual Spanish speakers from the capital during their visits for business, school, or pleasure. Due to their proximity to Brazil and frequent visits to Santana do Livramento for shopping or to visit with friends and relatives, contact with monolingual Brazilian Portuguese is also very intense. Added to steady exposure to national media from both Brazil and Uruguay television channels, Riverans are in daily contact with monolingual varieties of both Spanish and Portuguese, while maintaining their border dialects.

Figure 1.Map of Rivera and Montevideo
Figure 1.

3.Negative polarity tag questions as a discourse-pragmatic feature in Spanish and Portuguese

Tag questions have been studied in several languages, including Spanish (Domínguez Mujica 2005Domínguez Mujica, Carmen Luisa 2005 “Marcadores de (In)conclusión en el Español Hablado en Mérida-Venezuela. Boletín De Lingüística, 23: 3–22.Google Scholar; García Vizcaíno 2005García Vizcaíno, María José 2005 “El uso de los apéndices modalizadores¿ no? y¿ eh? en español peninsular.” In Selected Proceedings of the Second Workshop on Spanish Sociolinguistics ed. by Lotfi Sayahi and Maurice Westmoreland, 89–101. Somerville, MA: Cascadilla Proceedings Project.Google Scholar; Gómez-González 2014Gómez-González, María de los Ángeles 2014 “Canonical Tag Questions in English, Spanish and Portuguese: A Discourse-functional Study.” Languages in Contrast 14 (1): 93–126. CrossrefGoogle Scholar; Ocampo 2013Ocampo, Francisco 2013 “La Discursivización de Viste/Vio en Rioplatense.” In Perspectivas teóricas y experimentales sobre el español de la Argentina, ed. by Laura Colantoni and Celeste Rodríguez Louro, 487–505. Madrid: Iberoamericana. CrossrefGoogle Scholar; Rodríguez-Muñoz 2009Rodríguez-Muñoz, Francisco J. 2009 “Estudio sobre las funciones pragmadiscursivas de¿ no? y¿ eh? en el español hablado.” Revista de Lingüística Teórica y Aplicada 47 (1): 83–101.Google Scholar), Portuguese (Ferreira 1981Ferreira, Madalena 1981 “Tag Questions in Portuguese: Grammar and Intonation.” Phonetica 38 (5): 341–352. CrossrefGoogle Scholar; Gómez-González 2014Gómez-González, María de los Ángeles 2014 “Canonical Tag Questions in English, Spanish and Portuguese: A Discourse-functional Study.” Languages in Contrast 14 (1): 93–126. CrossrefGoogle Scholar; Gorski and Freitag 2006Gorski, Edair Maria and Raquel Meister Ko Freitag 2006 “Marcação e Comportamento Sociolinguístico de Marcadores Discursivos Interacionais de Base Verbal na Fala de Florianópolis.” In Variação e mudança linguística na Região Sul, ed. by Paulino Vandresen, 29–48. Pelotas, Brazil: EDUCAT.Google Scholar; Silva and Macedo 1992Silva, Giselle M. de Oliveira, and Alzira Tavares de Macedo 1992 “Discourse Markers in the Spoken Portuguese of Rio De Janeiro.” Language Variation and Change 4 (2): 235–49. CrossrefGoogle Scholar), Italian (Tomaselli and Gatt 2015Tomaselli, Maria Vittoria, and Albert Gatt 2015 “Italian Tag Questions and their Conversational Functions.” Journal of Pragmatics 84: 54–82. CrossrefGoogle Scholar), and Mandarin (Gao 2013Gao, Hua 2013 “Tag-Questions in Mandarin Chinese.” In Grammar in Cross-Linguistic Perspective The Syntax, Semantics, and Pragmatics of Japanese and Chinese ed. by Teruhiro Ishiguro and Kang Kwong Luke, 43–64. Bern, Switzerland: Peter Lang AG.Google Scholar), among others. Negative polarity tag questions consist of an anchor in the declarative, interrogative, explanative or imperative mood, followed by a tag in an interrogative form with negative polarity. In a cross-linguistic comparative study, Gómez-González (2014Gómez-González, María de los Ángeles 2014 “Canonical Tag Questions in English, Spanish and Portuguese: A Discourse-functional Study.” Languages in Contrast 14 (1): 93–126. CrossrefGoogle Scholar, 95) illustrates negative polarity tag questions in English, Spanish, and Portuguese showing highly similar structures (Figure 2)

Figure 2.Negative polarity tag questions in Spanish, Portuguese, and English (adapted from Gómez-González 2014Gómez-González, María de los Ángeles 2014 “Canonical Tag Questions in English, Spanish and Portuguese: A Discourse-functional Study.” Languages in Contrast 14 (1): 93–126. CrossrefGoogle Scholar)
Anchor   Tag   Language
Es suficiente,   ¿no?   Spanish
É suficiente,   não é?   Portuguese
That’s enough,   isn’t it?   English

While tags usually show reverse polarity, as illustrated in Figure 2, tags that preserve the same polarity as the anchor are also possible, although not as common. Canonical interrogative tags typically maintain the subject of the anchor, but other tags are invariant. Among invariant tags, one commonly finds both positive and negative polarity forms in both Spanish (¿verdad? ‘true?’and ¿no?’’no?’) and Portuguese (viu? ‘’did you see it’? and né? ‘innit?’).

The negative tag system in Portuguese presents structural differences from its Spanish counterpart. In European Portuguese, Ferreira (1981)Ferreira, Madalena 1981 “Tag Questions in Portuguese: Grammar and Intonation.” Phonetica 38 (5): 341–352. CrossrefGoogle Scholar distinguishes between a non-reduced invariant form não é? or a “negator + é” tag, from a variant “negator + V” tag in which the V stands for any verb in the anchor. In Brazilian Portuguese, a reduced invariant tag, né?, has been documented in addition to the non-reduced invariant form não é? that is found in European Portuguese. In de Oliveira e Silva and Macedo’s (1992)Silva, Giselle M. de Oliveira, and Alzira Tavares de Macedo 1992 “Discourse Markers in the Spoken Portuguese of Rio De Janeiro.” Language Variation and Change 4 (2): 235–49. CrossrefGoogle Scholar study of discourse markers in Rio de Janeiro, the authors found that , a reduced combination between não (‘no’) and é (‘is’), was by far the most frequent discourse marker. The use of né? in Brazilian Portuguese may be similar to the innovative use of use of innit in some varieties of British English, which originated from the canonical tag question isn’t it? (Anderson 2001; Pichler 2013 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar, 2016 2016 “Uncovering Discourse-Pragmatic Innovations: Innit in Multicultural London English.” In Discourse-Pragmatic Variation and Change in English: New Methods and Insights, ed by Heike Pichler, 59–84. Cambridge, UK: Cambridge University Press. CrossrefGoogle Scholar). This reduced variant is believed to be advancing among younger people in these varieties and is found not only in the right periphery of the sentence, the canonical position, but also in the right periphery of main clause anchors (Anderson 2001; Pichler 2016 2016 “Uncovering Discourse-Pragmatic Innovations: Innit in Multicultural London English.” In Discourse-Pragmatic Variation and Change in English: New Methods and Insights, ed by Heike Pichler, 59–84. Cambridge, UK: Cambridge University Press. CrossrefGoogle Scholar). Although there is no diachronic data available for the development of canonical tags in Portuguese that culminated in the use of reduced né? through phonetic attrition as it is the case with English innit (Pichler 2013 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar), the forms of negative polarity tag questions in both English and Portuguese present clear parallels, as illustrated in Figure 3. While the reduced form né? in Brazilian Portuguese presents obvious similarities with innit? in English, no monolingual varieties of Spanish have been reported to show a parallel development.

Figure 3.Reduced forms of negative polarity tag questions in Portuguese and English
Negator+ anchor V   Negator+ to be   Reduced form
Ele escreve, não escreve?   Ele escreve, não é?   Ele escreve, né?
He writes, doesn’t he?   He writes, isn’t it?   He writes, innit?

In addition to the forms of negative tag questions, their functions in discourse have also been analyzed. In a comparative study of tag questions in Spanish, Portuguese, and English, Gómez-González (2014)Gómez-González, María de los Ángeles 2014 “Canonical Tag Questions in English, Spanish and Portuguese: A Discourse-functional Study.” Languages in Contrast 14 (1): 93–126. CrossrefGoogle Scholar analyzes discourse functions of both positive and negative polarity tag questions by classifying them as addressee-centered, speaker-centered, and exchange-centered. Addressee-centered tag questions involve seeking information or confirmation from an interlocutor or mitigating a request to get an addressee to do something. Speaker-centered tags express a positive or negative stance of the speaker towards what they are saying, and exchange-centered tags are used to facilitate or align an exchange between interlocutors. Pichler (2013) 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar found similar functions of negative polarity tag questions in English. Instead of classifying tags as addressee-centered, speaker-centered, and exchange-centered, Pichler (2013) 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar categorizes the functions of negative polarity tag questions in British English as conducive if they intend to elicit a response from the interlocutor, or non-conducive if they do not intend to elicit a response. Similar to Gómez-Gonzalez’s (2014)Gómez-González, María de los Ángeles 2014 “Canonical Tag Questions in English, Spanish and Portuguese: A Discourse-functional Study.” Languages in Contrast 14 (1): 93–126. CrossrefGoogle Scholar category of addressee-centered tags, conducive negative polarity tag questions according to Pichler (2013) 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar can be used to seek the involvement of an interlocutor (involvement inducer) or to seek information (epistemic marker). Corresponding to Gómez-González’s (2014)Gómez-González, María de los Ángeles 2014 “Canonical Tag Questions in English, Spanish and Portuguese: A Discourse-functional Study.” Languages in Contrast 14 (1): 93–126. CrossrefGoogle Scholar categories of speaker-centered and exchange-centered tags, non-conducive negative polarity tag questions according to Pichler (2013) 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar can be used to express an attitude or stance towards an utterance (attitudinal stance marker), to signal alignment between interlocutors (alignment signal), or to mitigate (mitigation device). Crucially, Pichler (2013) 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar provides evidence that non-conducive negative polarity tag questions are a later development than conducive negative polarity tag questions and this semantic bleaching corresponds with the reduction of form, as seen in the use of the reduced form ‘innit?’

Freitag (2008)Freitag, Raquel Meister Ko 2008 “Marcadores Discursivos Interacionais na Fala de Itabaiana/SE.” Revista do Gelne, 10 (1): 21–32.Google Scholar explains a similar developmental process in Brazilian Portuguese in which addressee-centered markers that intend to elicit a response from an interlocutor develop new interpersonal and even textual or rhythm-making functions (2–3). Importantly, Freitag (2008)Freitag, Raquel Meister Ko 2008 “Marcadores Discursivos Interacionais na Fala de Itabaiana/SE.” Revista do Gelne, 10 (1): 21–32.Google Scholar classifies the reduced tag in Brazilian Portuguese in this last category, since it is free of referential meaning, extremely frequent in discourse, and is usually used to “keep and maintain the rhythm of the speaker’s turn” (3). Thus, the reduced form né? in Brazilian Portuguese shows evidence of semantic bleaching, increased frequency, and phonetic attrition when compared to canonical non-reduced negative tags, all steps towards grammaticalization, and similar to what has been found for the development of English ‘innit?’ (Andersen 2001Andersen, Gisle 2001 “Pragmatic Markers and Sociolinguistic Variation a Relevance-theoretic Approach to the Language of Adolescents.” Pragmatics & Beyond 84. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing.Google Scholar; Pichler 2013 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar, 2016 2016 “Uncovering Discourse-Pragmatic Innovations: Innit in Multicultural London English.” In Discourse-Pragmatic Variation and Change in English: New Methods and Insights, ed by Heike Pichler, 59–84. Cambridge, UK: Cambridge University Press. CrossrefGoogle Scholar).

Although Spanish does not present a reduced form from a canonical negative tag, the invariant use of ¿no? has also been found to fulfill a similar discourse device function of bracketing units of talk similar to in Brazilian Portuguese. In a qualitative study of the functions of the non-interrogative no in Mexican Spanish, Vásquez Carranza (2017)Vázquez Carranza, Ariel 2017 “Some Uses of ‘no’ in Spanish Talk-in-Interactions.” International Review of Pragmatics 9, (2): 224–247. CrossrefGoogle Scholar finds a few uses of a discourse-oriented ¿no?. Similar to né?, this ¿no? does not intend to elicit a response from an interlocutor but is used to structure or organize talk. Although the author offers a few examples that illustrate the use of ¿no? to fulfill a discourse device function, the frequency of this function in Spanish is not clear due to the lack of quantitative analysis.

In summary, we highlight two important differences between Spanish and Portuguese in the use of negative tags. First, while Portuguese presents both canonical “negator + V” forms and reduced forms that originated from the “negator + é”, Spanish presents a much more reduced system comprised solely of the isolated and invariant ¿no? particle for negative polarity tags. To our knowledge, the only use of the invariant “negator + es” (¿no es?) in Spanish was found in northern Uruguay and attributed to Portuguese influence (Elizaincín 1995Elizaincín, Adolfo 1995 “Personal Pronouns for Inanimate Entities in Uruguayan Spanish in Contact with Portuguese.” In Spanish in Four Continents: Studies in Language Contact and Bilingualism, ed. by Carmen Silva-Corvalán, 117–131. Washington, D.C.: Georgetown University Press.Google Scholar). Secondly, in terms of function, while in Brazilian Portuguese mainly fulfills a non-conducive discourse device function that brackets units of discourse (Freitag 2008Freitag, Raquel Meister Ko 2008 “Marcadores Discursivos Interacionais na Fala de Itabaiana/SE.” Revista do Gelne, 10 (1): 21–32.Google Scholar), there is no quantitative study available in Spanish that points to such a preference for ¿no?.

The contrast in both form and function of negative polarity tag questions in Spanish and Portuguese allows for the exploration of the permeability of negative polarity tag systems among bilinguals. Here we ask, first, whether the negative polarity tag question system in Spanish found among monolingual Spanish speakers from the Uruguayan capital, Montevideo, deviates from the one found among Portuguese-Spanish bilinguals due to language contact. We then ask whether any deviation could be explained by patterns found in the Portuguese variety spoken by the same bilinguals. The ultimate goal of this study is to test whether the use of negative polarity tag questions substantiates the assumption of discourse-pragmatic features’ high permeability in language contact situations, especially between cognate systems, or if the systems remain independent, despite long term contact. We will first look at forms of negative polarity tag questions in both Spanish and Portuguese, since we will need to first establish the variants’ underlying structural similarities and differences (Pichler 2010Pichler, Heike 2010 “Methods in Discourse Variation Analysis: Reflections on the Way Forward 1.” Journal of Sociolinguistics 14 (5): 581–608. CrossrefGoogle Scholar, 590). Once the forms are defined, the functions fulfilled by these forms are analyzed and compared. The analysis of the functions fulfilled by negative tags in monolingual Spanish will set the basis for comparison with the variety spoken by bilinguals and fulfill the need to assess the extent that these forms fulfill conducive and non-conducive functions. By comparing the distribution of both forms and functions of negative polarity tag in Spanish and Portuguese, we hope to shed light on the permeability of cognate discourse-pragmatic systems in prolonged contact.

4.Method

When assessing the permeability of languages in contact, it is important to subject similar data sets of different languages to the same analysis (Poplack and Levey 2010Poplack, Shana, and Stephen Levey 2010 “Contact-induced Grammatical Change: A Cautionary Tale.” In Language and Space: An International Handbook of Linguistic Variation 1, ed. by Jürgen Erich Schmidt and Peter Auer, 391–419. New York, NY: Walter de Gruyter.Google Scholar; Tagliamonte 2013Tagliamonte, Sali A. 2013 “Comparative Sociolinguistics.” In The Handbook of Language Variation and Change Vol. 129, ed. by Jack K. Chambers and Natalie Schilling: 128–156. New York, NY: John Wiley & Sons. CrossrefGoogle Scholar). The bilingual corpus includes a set of interviews in Spanish and a set of interviews in Portuguese with the same speakers. These interviews followed the traditional format of a sociolinguistic interview and lasted approximately one hour. After collecting demographic information from the participants, the interviewer brought up topics that would elicit spontaneous talk. The participants were contacted through a friend of a friend (snow ball technique), and approached first in Spanish, the national language preferred among strangers. After the interview in Spanish took place, the interviewer asked permission to come back another time to continue the conversation. Having established a relationship with the participants, the second interview was carried out in Portuguese, the language preferred among acquaintances. The interviewer was bilingual and fluent in the local dialects. Both interviews were recorded and later transcribed in their entirety. All speakers were raised bilingual by bilingual families, used both languages in their everyday lives, and presented no difficulties in carrying out an hour-long conversation in either language.

Because discourse-pragmatic variation has “extreme context-sensitivity”, it is fundamental for comparisons to be carried out between similar genres in order to assure “cross-corpora comparability and generalizability” since several factors, including role of interlocutors and discourse type, may impact on the frequency and function of discourse-pragmatic features (Pichler 2010Pichler, Heike 2010 “Methods in Discourse Variation Analysis: Reflections on the Way Forward 1.” Journal of Sociolinguistics 14 (5): 581–608. CrossrefGoogle Scholar, 585). We follow Pichler’s suggestion that these problems can be remediated if similar corpora are compared (586), and compared 18 sociolinguistic interviews with Spanish-Portuguese bilinguals from Rivera with 18 sociolinguistic interviews with Spanish-speaking monolinguals from Montevideo. The monolingual corpus was extracted from the PRESEEA (http://​preseea​.linguas​.net/), while the interviews in Spanish with bilinguals were extracted from a large corpus of interviews conducted in Rivera by Carvalho. Both corpora present similar sociolinguistic stratification regarding sex and age groups.

Table 1.Rivera and Montevideo corpora including sex and age groups
Spanish-Portuguese bilinguals from Rivera, Uruguay Spanish monolinguals from Montevideo, Uruguay
Sex 9 women and 9 men 9 women and 9 men
Age groups Group 1, 16–29: 6 Group 2, 30–49: 6
Group 3, 50+: 6 Group 1, 16–29: 5
Group 2, 30–49: 5 Group 3, 50+: 8

230 tokens of negative polarity tag questions were extracted from the Montevideo corpus and compared to 362 tokens used by Spanish-Portuguese bilinguals in Rivera during their interviews in Spanish. In addition, negative polarity tag questions in Portuguese used by the same participants were analyzed as well, based on 382 tokens. Once the tokens were extracted, their forms and functions were analyzed and compared across these three data sets. First, we identified and quantified all the possible variants of this variable in these three varieties. The Montevideo corpus was analyzed so that the bilingual dialect could be compared to what Poplack and Levey (2010)Poplack, Shana, and Stephen Levey 2010 “Contact-induced Grammatical Change: A Cautionary Tale.” In Language and Space: An International Handbook of Linguistic Variation 1, ed. by Jürgen Erich Schmidt and Peter Auer, 391–419. New York, NY: Walter de Gruyter.Google Scholar call an “appropriate reference.” Once the source dialect was examined and compared with the bilingual variety, we turned our attention to the inventory of negative polarity tag questions in Uruguayan Portuguese to compare and contrast the two pragmatic systems that coexist in the community and assess the influence that they may have on each other.11.Since no corpus of Sourthern Brazilian Portuguese is available, it is not possible to extend this comparison to include a monolingual variety of Portuguese.

5.Analysis

5.1Forms of negative polarity tag questions

All of the forms of negative polarity tag questions in Montevideo Spanish and Rivera Spanish were identified and extracted from the three corpora for analysis. Results are shown in Table 2.

Table 2.Forms of negative polarity tag questions in Spanish among Spanish monolinguals in Montevideo and among Spanish-Portuguese bilinguals in Rivera
Border Spanish forms Total
¿no? 318/362 (87.8%)
né? 20/362 (5.5%)
¿no es? 15/362 (4.1%)
¿no é?  3/362 (0.8%)
não é?  2/362 (0.6%)
¿no es así?  2/362 (0.6%)
¿no fue?  1/362 (0.3%)
¿no ves?  1/362 (0.3%)
Total 362/362 (100%)
Montevideo Spanish Forms Total
¿no? 230/230 (100%)

As seen in Table 2, in Montevideo Spanish, the non-contact variety, the only form of negative polarity tag questions was ¿no? (100%), confirming Gómez-González’s (2014)Gómez-González, María de los Ángeles 2014 “Canonical Tag Questions in English, Spanish and Portuguese: A Discourse-functional Study.” Languages in Contrast 14 (1): 93–126. CrossrefGoogle Scholar results that showed the use of ¿no? as the only invariant negative polarity tag question in monolingual Spanish. In border Spanish, the contact variety, ¿no? was also the overwhelmingly preferred variant (87.8%). However, unlike the monolingual variety, other forms of negative polarity tag questions in Rivera Spanish included a direct borrowing of né? from Portuguese (5.5%), followed by ¿no es? (4.1%), a calque from the Portuguese tag não é?. Alternative forms were used scarcely as well (¿no é?, não é?, ¿no fue?, etc.), but accounted for for less than 3% of the data.

Since inter-individual variation is commonly found in the use of discourse markers in general (Sankoff et al. 1997Sankoff, Gillian, Pierrette Thibault, Naomi Nagy, Hélène Blondeau, Marie-Odile Fonollosa, and Lucie Gagnon 1997 “Variation in the Use of Discourse Markers in a Language Contact Situation.” Language Variation and Change 9 (2): 191–217. CrossrefGoogle Scholar, among others) and in the use of tag questions in particular (Ferreira 1981Ferreira, Madalena 1981 “Tag Questions in Portuguese: Grammar and Intonation.” Phonetica 38 (5): 341–352. CrossrefGoogle Scholar), it is important to consider individual behavior behind the frequency of the forms. Table 3 illustrates the distribution of the forms of negative polarity tag questions in Spanish across individuals in Rivera.

Table 3.Forms of negative polarity tag questions in Border Spanish among Spanish-Portuguese bilinguals in Rivera (per participant)
Participant Total
A B C D E F G H I J K L M N O P Q R
¿no? 11 4 1 27 19 1 4 3 16 25 70 32 39 1 12 40 5 9 318/362 (87.8%)
né?  0 0 0  1 18 0 0 0  0  0  0  0  0 0  0  0 1 0 20/362 (5.5%)
¿no es?  1 0 0  1  1 1 0 0  0  2  2  0  2 0  4  1 0 0 15/362 (4.1%)
¿no é?  0 0 0  0  1 0 0 0  0  1  0  0  0 0  1  0 0 0  3/362 (0.8%)
não é?  0 0 0  0  2 0 0 0  0  0  0  0  0 0  0  0 0 0  2/362 (0.6%)
¿no es así?  0 0 0  0  0 0 0 0  0  0  0  0  0 0  2  0 0 0  2/362 (0.6%)
¿no fue?  0 0 0  0  0 0 0 0  0  0  0  0  0 0  1  0 0 0  1/362 (0.3%)
¿no ves?  0 0 0  0  1 0 0 0  0  0  0  0  0 0  0  0 0 0  1/362 (0.3%)
Total 12 4 1 29 42 1 4 3 16 28 72 32 41 1 20 41 6 9 362/362 (100%)

The distribution of these forms in Table 3 clearly shows that the only negative polarity tag question that is used by everyone during the interview in Spanish is the canonical ¿no?. The second form that is used by several individuals is the calque ¿no es? from Portuguese não é? (8 individuals out of 18, or 40.4% of the sample). Although ¿no es? was more widespread among participants, it only accounted for 4.1% (15/362) of negative polarity tag questions in the corpus. Slightly more frequent than this calque was the use of the direct borrowing form Portuguese né?; however, this form was only used in Spanish by three participants. One participant in the corpus (Participant E) produced 18 of the 20 tokens. All of the other forms were used only once or twice by no more than 5 participants. Taken together, these results show that in terms of the forms of negative polarity tag questions, the Spanish of these bilinguals highly coincides with the patterns found in Montevideo. The calque ¿no es? may constitute a feature of the community grammar that departs from the monolingual norm, as previously claimed by Elizaincín (1995)Elizaincín, Adolfo 1995 “Personal Pronouns for Inanimate Entities in Uruguayan Spanish in Contact with Portuguese.” In Spanish in Four Continents: Studies in Language Contact and Bilingualism, ed. by Carmen Silva-Corvalán, 117–131. Washington, D.C.: Georgetown University Press.Google Scholar. All of the other forms are scarce and should be considered idiosyncratic.

In order to assess the extent to which the bilingual behavior found in the differential use of negative tag question forms in Spanish is due to direct influence of Portuguese, it is important to look at the use of this variable by the same bilinguals in Portuguese. This information is presented in Table 4.

Table 4.Forms of negative tag questions in Border Portuguese among Spanish-Portuguese bilinguals in Rivera
Forms Total
né? 182/382 (47.6%)
não? 126/382 (33.0%)
não é?  51/382 (13.4%)
¿no? 23/382 (6.0%)
Total 382/382 (100%)

In Uruguayan Portuguese, né? was the most frequent negative polarity tag question (47.6%;), a tendency also documented in previous studies of Brazilian Portuguese (i.e. Gorski and Freitag 2006Gorski, Edair Maria and Raquel Meister Ko Freitag 2006 “Marcação e Comportamento Sociolinguístico de Marcadores Discursivos Interacionais de Base Verbal na Fala de Florianópolis.” In Variação e mudança linguística na Região Sul, ed. by Paulino Vandresen, 29–48. Pelotas, Brazil: EDUCAT.Google Scholar). Other negative polarity tag questions in the corpus included não? (33.0%; 126/382) and não é? (13.4%; 51/382), which both coincide with the Portuguese monolingual norm. The only contact-induced form found in Portuguese was the use of a direct borrowing from Spanish, ¿no? (6.0%; 23/382).

Let’s now turn to individual behavior in the use of negative polarity tag questions in Portuguese in Table 5 in order to see if the use of Spanish ¿no? in Uruguayan Portuguese is widespread or idiosyncratic.

Table 5.Forms of negative polarity tag questions in Border Portuguese among Spanish-Portuguese bilinguals in Rivera (per participant)
Participant Total
A B C D E F G H I J K L M N O P Q R
né?  4  9 3 10 48 3 3 11 0  5 30  6 18 2  4 12 14 0 182/382 (47.6%)
não?  8  3 0 14  0 3 3 11 2 12 19  7 19 0  6 18  0 1 126/382 (33.0%)
não é?  2  0 0  3  1 0 0  0 0  3  1  3  6 0  7 23  0 2  51/382 (13.4%)
¿no?  1  1 0  3  0 3 0  0 0  2  3  1  2 2  2  2  0 1 23/382 (6.0%)
Total 15 13 3 30 49 9 6 22 2 22 53 17 45 4 19 55 14 4 382/382 (100%)

The results in Table 5 show that while né? is the only form that is used by all individuals, mirroring closely the Brazilian Portuguese system. All of the other forms are also used by the majority of the participants, including the borrowing of Spanish ¿no?, used by 12 out of 18 participants. This suggests that the use of Spanish ¿no? in Uruguayan Portuguese is more generalized across participants than the borrowing of Portuguese né? in Rivera Spanish, as illustrated in Figure 5. This discrepancy follows the tendency among minority language speakers to incorporate borrowings from the majority language in general, and discourse makers in particular, as found in multiple studies (Brody 1987Brody, Jill 1987 “Particles Borrowed From Spanish As Discourse Markers in Mayan Languages.” Anthropological Linguistics 29 (4): 507–521.Google Scholar; Flores-Ferrán 2014Flores-Ferrán, Nydia 2014 “So Pues Entonces: An Examination of Bilingual Discourse Markers in Spanish Oral Narratives of Personal Experience of New York City-born Puerto Ricans.” Sociolinguistic Studies 8 (1): 57–79. CrossrefGoogle Scholar; Salmons 1990Salmons, Joe 1990 “Bilingual Discourse Marking: Code Switching, Borrowing, and Convergence in Some German-American Dialects.” Linguistics 28 (3): 453–480. CrossrefGoogle Scholar; Torres 2002Torres, Lourdes 2002 “Bilingual Discourse Markers in Puerto Rican Spanish.” Language in Society 31 (1): 65–83. CrossrefGoogle Scholar; Torres and Potowski 2008Torres, Lourdes, and Kim Potowski 2008 “A Comparative Study of Bilingual Discourse Markers in Chicago Mexican, Puerto Rican, and MexiRican Spanish.” International Journal of Bilingualism 12 (4): 263–279. CrossrefGoogle Scholar). In Rivera, it is also the case that while Spanish, the majority language, lends ¿no? to Portuguese, the minority language, in general, the borrowing of Portuguese né? in Spanish is much more limited and idiosyncratic, as shown in Figure 4.

Figure 4.Non-generalized borrowing of né? in border Spanish and generalized borrowing of ¿no? in border Portuguese in Rivera
Figure 4.

In summary, there are strong continuities between Rivera Spanish and Montevideo Spanish in terms of the forms of negative polarity tag questions. Contact with Portuguese is only seen in the few but generalized instances of ¿no es?, a calque from Portuguese não é?. The use of discursive calques in bilingual situations was also detected by Sankoff et al. (1997)Sankoff, Gillian, Pierrette Thibault, Naomi Nagy, Hélène Blondeau, Marie-Odile Fonollosa, and Lucie Gagnon 1997 “Variation in the Use of Discourse Markers in a Language Contact Situation.” Language Variation and Change 9 (2): 191–217. CrossrefGoogle Scholar, who found generalized use of French ‘comme’, a calque from English ‘like’, among French-English bilinguals in Canada. The presence of the direct borrowing of né? from Portuguese in the interviews in Spanish was also detected but much less generalized and only present in the speech of three participants, despite its massive presence in the Portuguese speech of the same bilinguals. The Portuguese variety spoken by bilinguals also showed similarities towards Brazilian Portuguese, with the exception of the use of ¿no?, a discourse marker directly borrowed from Spanish, the majority language, resembling what has been found, for example, in the use of the English discourse marker ‘so’ in by Spanish-English bilinguals in the United States (Aaron 2004Aaron, Jessi 2004 ““So Respetamos Un Tradicion Del Uno Al Otro.” So and Entonces in New Mexican Bilingual Discourse.” Spanish in Context 1 (2): 161–79. CrossrefGoogle Scholar). Therefore, we find very little evidence of the pragmatic borrowing (Andersen 2014 2014 “Pragmatic Borrowing.” Journal of Pragmatics 67: 17–33. CrossrefGoogle Scholar) of the forms of negative polarity tag questions in a situation of intense language contact. Distinct from several previous studies of discourse-pragmatic features, there is no evidence of the convergence of forms from two linguistic systems (Dajko and Carmichael 2014Dajko, Nathalie, and Katie Carmichael 2014 “But Qui C’est La Différence ? Discourse Markers in Louisiana French: The Case of but vs. mais.” Language in Society 43 (2): 159–83. CrossrefGoogle Scholar; de Rooij 2000de Rooij, Vincent 2000 “French Discourse Markers in Shaba Swahili Conversations.” International Journal of Bilingualism 4 (4): 447–67. CrossrefGoogle Scholar; Hlavac 2006Hlavac, Jim 2006 “Bilingual Discourse Markers: Evidence from Croatian-English Code-switching.” Journal of Pragmatics 38 (11): 1870–1900. CrossrefGoogle Scholar; Maschler 2009Maschler, Yael 2009Metalanguage in Interaction: Hebrew Discourse Markers. Vol. 181. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar; Matras 1998Matras, Yaron 1998 “Utterance Modifiers and Universals of Grammatical Borrowing.” Linguistics 36 (2): 281–332. CrossrefGoogle Scholar, 2000 2000 “Fusion and the Cognitive Basis for Bilingual Discourse Markers.” International Journal of Bilingualism 4 (4): 505–528. CrossrefGoogle Scholar).

5.2Functions of negative polarity tag questions

After analyzing the forms of negative polarity tag questions in border Spanish and comparing them to the monolingual variety of Spanish in Uruguay and to Uruguayan Portuguese, we turn to the analysis of the discourse functions. Our question here is if the tag question system of the Spanish variety spoken by bilinguals compares to the one spoken by monolinguals in Montevideo. Since the negative tag né? in Brazilian Portuguese is used mainly as a discourse device that is highly non-conducive and is related more to structuring discourse rather than seeking to elicit a response from an interlocutor (Gorski and Freitag 2006Gorski, Edair Maria and Raquel Meister Ko Freitag 2006 “Marcação e Comportamento Sociolinguístico de Marcadores Discursivos Interacionais de Base Verbal na Fala de Florianópolis.” In Variação e mudança linguística na Região Sul, ed. by Paulino Vandresen, 29–48. Pelotas, Brazil: EDUCAT.Google Scholar), we ask if any divergence from the Spanish monolingual variety, if found, can be attributed to Portuguese influence.

In addition to the discourse device function, our analysis includes six other discourse functions identified for negative polarity tag questions in Pichler’s (2013) 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar variationist analysis of negative polarity tag questions in British English: involvement inducer, epistemic marker, attitudinal stance marker, alignment signal, and mitigation. We have chosen Pichler’s (2013) 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar classification of functions because our similar methodologies, especially our focus solely on negative polarity tag questions, permit cross-linguistic comparisons. Moreover, similar functions have been documented for tag questions in both Spanish and Portuguese by Gómez-González (2014)Gómez-González, María de los Ángeles 2014 “Canonical Tag Questions in English, Spanish and Portuguese: A Discourse-functional Study.” Languages in Contrast 14 (1): 93–126. CrossrefGoogle Scholar. Following Pichler (2013) 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar, the functions of negative polarity tag questions were further classified as conducive, if they elicited or intended to elicit a response from the interlocutor, or non-conducive, if they did not intend to elicit a response. Non-conducive functions allow negative polarity tag questions to be used to mitigate what was said, bring focus to what was said, or only bracket discourse as seen with the discourse device function.

5.2.1The involvement inducer function

The involvement inducer function “seeks corroboration of propositions” and “secures addressees’ involvement in discourse” (Pichler 2013 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar, 270). In (1), the participant uses the negative polarity tag question ¿no? to seek corroboration that the two sides of the border are similar but not the same. This function is conducive because it elicits a response from the interlocutor.

(1)

I:
Parecida pero no igual. Porque ellos tienen la manera de ellos y nos tenemos la nuestra, ¿no? Ellos se copian mucho la manera de ser, ¿no? A mi manera de pensar, ¿no?
E:
Claro, claro. Es lo que quiero saber, tú manera de pensar (Rivera, 12A)
I:
Similar but not the same. Because they have their own way and we have our own, no? They copy our way of being a lot, no? In my way of thinking, no?
E:
Of course, of course. This is what I want to know, your way of thinking.

5.2.2The epistemic marker function

The epistemic marker function “signals speakers’ uncertainty” regarding preceding propositions and “seeks verification” (Pichler 2013 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar, 269). In (2), the participant uses ¿no? to signal uncertainty about the name of the city on the other side of the bridge. This function is also conducive because it elicits a response.

(2)

I:
Tá el puente. Hay otro puente internacional ahí. Del otro lado es Cuareina, ¿no?
E:
No sé (Rivera, 58A)
I:
The bridge is there. There is another international bridge there. On the other side is Cuareina, no?
E:
I don’t know

5.2.3The mitigation device

The mitigation device function “mitigates face-threat of dismissals, disagreements, and ironic or sarcastic remarks” (Pichler 2013 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar, 269). This function can be conducive or non-conducive. In (3), the participant uses ¿no? to mitigate his remark that he reads very little. In this case, the mitigation device function is conducive because it elicits a response.

(3)

E:
¿Y te gusta leer el diario?
I:
Leo muy poco. No compro el diario. Es un defecto mío, ¿no?
E:
Defecto. Y cuando lees el diario, lees el diario uruguayo,brasileño o norteño? (Rivera 8A)
E:
And do you like to read the newspaper?
I:
I read very little. I don’t buy the newspaper. It a personal defect of mine, no?
E:
defect. And when you read the newspaper, do you read the Uruguayan, Brazilian, or Northern newspaper?

5.2.4The attitudinal stance

The attitudinal stance marker function “foregrounds and emphasizes preceding propositions” (Pichler 2013 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar, 270). In (4), ¿no? emphasizes the participant’s preceding opinion that it is good to speak Spanish and Portuguese well. This function is non-conducive because it does not elicit a response from the interlocutor.

(4)

Digo. Es lindo, es bueno saber hablar el castellano bien y el portugués, ¿no? Es lindo saber otro idioma. (Rivera, 16A)

I say. It is beautiful, It is good to know Spanish well and Portuguese, no? It is good to know another language.

5.2.5The alignment signal

The alignment signal function “signals alignment with previous speakers and involvement in interaction” (Pichler 2013 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar, 269), as shown in (5). This function is also non-conducive because it does not elicit a response.

(5)

I:
Dijeron “Bueno, XXX, nosotros esperamos tu contestación. Éste. Es lamentable. Nosotros vemos muchas cualidades dentro de ti, muchas. Muchas cualidades. Pero nosotros no te podemos decir lo que tenés que hacer. Vos tenés que tomar una decisión. Éste. Solo.
E:
Solo, claro.
I:
Solo, ¿no? Por voluntad propia. (Rivera, 53A)
I:
They said: “Well, XXX, we wait your answer. This is lamentable. We see many qualities in you, many. Many qualities. But we cannot tell you what you have to do. You have to make a decision. On your own.
E:
On your own, of course.
I:
On my own, no? Willingly.

5.2.6The discursive device

Lastly, the discursive device function brackets elements of discourse (Gorski and Freitag 2006Gorski, Edair Maria and Raquel Meister Ko Freitag 2006 “Marcação e Comportamento Sociolinguístico de Marcadores Discursivos Interacionais de Base Verbal na Fala de Florianópolis.” In Variação e mudança linguística na Região Sul, ed. by Paulino Vandresen, 29–48. Pelotas, Brazil: EDUCAT.Google Scholar), as shown in (6). This function is also non-conducive because it does not elicit a response, and it is the function that shows the most advanced stage of grammaticalization, since any connection with the interlocutor is bleached and the particle is completely orientated towards textual organization.

(6)

Y yo juntamente con ellos aprendí a trabajar con los demás. No digo que me acostumbré porque fue un momento muy corto, ¿no? Pero me agarré como un costumbre, ¿no? A trabajar y hacer algo por los demás, ¿no? Y siempre me gustó esa vida, siempre me gustó (Rivera, 53A)

And I along with them learned how to work with the others. I’m not saying that I got used to it because it was a very short time, no? But it became a habit, no? To work and do something for others, no? And I always liked that life, I always liked it.

In summary, the conducive functions include the epistemic marker and involvement inducer while the non-conducive functions include the attitudinal stance marker, alignment signal, and discursive device functions. The mitigation device function can be either conducive or non-conducive, depending on if it is used to elicit a response. After classifying the negative polarity tag questions in the present corpora according to the six functions previously outlined, these functions were analyzed quantitatively. The functions of negative polarity tag questions in border Spanish, the contact variety, are presented in Table 6.

Table 6.Functions of negative polarity tag questions in Border Spanish among Spanish-Portuguese bilinguals in Rivera
Function Total
Conducive functions Involvement inducer  60/366 (16.4%)
Epistemic marker  22/366 (10.4%)
Mitigation device  2/366 (0.5%)
Total 84/366 (23.0%)
Non-conducive functions Attitudinal stance marker 164/366 (44.8%)
Discursive device 108/366 (29.5%)
Mitigation device  8/366 (4.7%)
Alignment signal  2/366 (0.5%)
Total 282/366 (77.0%)

In border Spanish, non-conducive functions are strongly favored over conducive functions, with 77.0% of negative polarity tag questions used to fulfill non-conducive functions. In the conducive category, the most frequently fulfilled function is involvement inducer, followed by the epistemic marker. In the non-conducive category, the most frequent fulfilled function was the attitudinal stance marker, followed by the discursive device. Following the premises of comparative sociolinguistics (Poplack and Levey 2010Poplack, Shana, and Stephen Levey 2010 “Contact-induced Grammatical Change: A Cautionary Tale.” In Language and Space: An International Handbook of Linguistic Variation 1, ed. by Jürgen Erich Schmidt and Peter Auer, 391–419. New York, NY: Walter de Gruyter.Google Scholar; Tagliamonte 2013Tagliamonte, Sali A. 2013 “Comparative Sociolinguistics.” In The Handbook of Language Variation and Change Vol. 129, ed. by Jack K. Chambers and Natalie Schilling: 128–156. New York, NY: John Wiley & Sons. CrossrefGoogle Scholar), we now turn to results of the same analysis applied to the monolingual corpus, presented in Table 7.

Table 7.Functions of negative polarity tag questions in Spanish among Spanish monolinguals in Montevideo
Function Total
Conducive functions Involvement inducer 108/237 (45.6%)
Epistemic marker   2/237 (0.08%)
Mitigation device   2/237 (0.08%)
Total 112/237 (47.3%)
Non-conducive functions Attitudinal stance marker  70/237 (29.5%)
Alignment signal  36/237 (15.2%)
Discursive device 11/237 (4.6%)
Mitigation device  8/237 (3.4%)
Total 125/237 (52.7%)

In the non-contact variety of Spanish in Montevideo, there is approximately an even distribution between conducive and non-conducive functions, with 52.7% of negative polarity tag questions used to fulfill non-conducive functions and 47.3% tags that fulfill functions that are conducive. The function within the conducive category that is preferred by far is the involvement inducer, followed by only 4 tokens that fulfill the other functions. The functions within the non-conducive category are more diverse. First, the attitudinal stance marker is favored, followed by the alignment signal. Both functions account for the vast majority of non-conducive realizations. The discourse device and mitigation device function appear rarely, accounting for 4.6% and 3.4% of the data, respectively.

When comparing the functions of negative polarity tag questions in the contact and non-contact varieties of Spanish, we see an important distinction. Diverging from the non-contact variety in Montevideo, non-conducive functions are strongly favored over conducive functions in the contact variety, with 77.0% of negative polarity tag questions fulfilling a non-conducive function in Rivera Spanish. Within the conducive category, negative polarity tag questions in the contact variety in Rivera show the same distribution found in the non-contact variety in Montevideo, with the great majority of the tokens fulfilling the involvement inducer function. Within the non-conducive category, however, while the most frequent function in both varieties is the attitudinal stance marker, the second most frequent function in the contact variety in Rivera is the discourse device. This function accounts for nearly 30% of negative polarity tag questions among bilinguals in Rivera (as opposed to only 4.6% of negative polarity tag questions among monolinguals in Montevideo). Since conducivity is viewed as an important factor in the grammaticalization of negative polarity tag questions, these results suggest that negative polarity tag questions may not only be somewhat more advanced in terms of grammaticalization in the Spanish spoken by bilinguals, but that this innovation may have been brought to the dialect through the discourse device function, the one found in the Brazilian Portuguese né?.

In order to attest the possibility that the higher use of non-conducive functions of negative tags in border Spanish could have been influenced by Portuguese, it is essential to analyze the functions of this pragmatic marker in the Portuguese spoken by these bilinguals so that we can understand the negative tag system in the language that has the potential to influence their Spanish. Recall that Freitag’s (2008)Freitag, Raquel Meister Ko 2008 “Marcadores Discursivos Interacionais na Fala de Itabaiana/SE.” Revista do Gelne, 10 (1): 21–32.Google Scholar study of discourse markers in Brazilian Portuguese pointed to a tendency of the tag né? to be very frequent (which was attested to be the case in Uruguayan Portuguese as well), and to be used mostly as a discourse device, a non-conducive function. We turn to the analysis of functions of the tags in Uruguayan Portuguese to verify, first, if this dialect shows the same patterns found in Brazilian Portuguese by Freitag (2008)Freitag, Raquel Meister Ko 2008 “Marcadores Discursivos Interacionais na Fala de Itabaiana/SE.” Revista do Gelne, 10 (1): 21–32.Google Scholar, and second, if it can help us understand the discrepancy found between Rivera Spanish and Montevideo Spanish. Table 8 presents the functions of negative polarity tag questions in Rivera Portuguese.

Table 8.Functions of negative polarity tag questions in Border Portuguese among Spanish-Portuguese bilinguals in Rivera
Function Total
Conducive functions Involvement inducer 40/498 (8.0%)
Epistemic marker 30/498 (6.0%)
Mitigation device  8/498 (1.6%)
Total 78/498 (15.7%)
Non-conducive functions Attitudinal stance marker 229/498 (46.0%)
Discursive device 122/498 (24.5%)
Mitigation device  54/498 (10.8%)
Alignment signal 15/498 (3.0%)
Total 420/498 (84.3%)

It is clear from the results displayed on Table 8 that non-conducive functions are even more strongly favored over conducive functions (84.3%) in the Portuguese variety spoken by Spanish-Portuguese bilinguals in Rivera, significantly more than Montevideo Spanish (47.3%) and also more than border Spanish (77%).

Taking all three varieties together, while the most frequently fulfilled function in Montevideo Spanish is the involvement inducer function, the attitudinal stance function is the most frequent in both Rivera Spanish and Rivera Portuguese. Kimps, Davidse and Cornillie (2014)Kimps, Ditte, Kristin Davidse, and Bert Cornillie 2014 “The Speech Functions of Tag Questions and their Properties. A Comparison of their Distribution in COLT and LLC.” Corpus Interrogation and Grammatical Patterns, 321–50. Amsterdam, Netherlands: John Benjamins. CrossrefGoogle Scholar remark that only a minority (20%) of tag questions in their study about English tags are used as “real information seeking questions” (75). This was also the case in the present study, in which the epistemic marker function accounted for only 10% of negative polarity tag questions in border Spanish, 6% in border Portuguese, and less than 1% in Montevideo Spanish.

Interestingly, and probably due to pre-contact typological coincidence, the conducive functions fulfilled by the negative tags in Portuguese are distributed in the same order as the one attested for both bilingual and monolingual Spanish varieties: involvement inducer, epistemic marker, and mitigation device, as shown in Figure 5.

Figure 5.Continuities of conducive functions in Montevideo Spanish, border Spanish, and border Portuguese
Figure 5.

Turning to the non-conducive functions, we see the same order for border Spanish and border Portuguese (attitudinal stance marker, discourse device, mitigation device, and alignment signal), but the alignment signal function ranks second after the attitudinal stance marker instead of the discourse device in Montevideo Spanish, as shown in Figure 6.

Figure 6.Non-conducive functions in Montevideo Spanish, border Spanish and border Portuguese
Figure 6.

The fact that the discourse device function follows the attitudinal stance marker function in both Rivera Spanish and Rivera Portuguese, but is relatively infrequent in Montevideo Spanish, brings evidence to the fact that this non-conducive function found to be common in Brazilian Portuguese (Freitag 2008Freitag, Raquel Meister Ko 2008 “Marcadores Discursivos Interacionais na Fala de Itabaiana/SE.” Revista do Gelne, 10 (1): 21–32.Google Scholar), is frequently fulfilled in border Spanish, presenting an important difference from the monolingual counterpart spoken in Montevideo.

In order to confirm that tag questions are in general more non-conducive in border Portuguese than border Spanish, mirroring the Brazilian Portuguese system, and that tag questions in border Spanish are more non-conducive than Montevideo Spanish due to the influence of the Portuguese system, we fit a generalized linear mixed-effects model with non-conducivity as the criterion and linguistic variety (border Spanish, border Portuguese, Montevideo Spanish) as the fixed effect. No factors other than linguistic variety were included in the analysis, as the sample was too small to be representative of social sub-groups. Since non-conducivity is categorical, the error term was binomially distributed with a logit linking function. Participants were given random intercepts. We tested for a main effect of community via nested model comparisons. All models were fit in R (R Core Team 2013R Core Team 2013R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, http://​www​.R​-project​.org) using the lme4 package (Bates, Mächler, Bolker and Walker 2015Bates, Douglas, Martin Mächler, Ben Bolker, and Steve Walker 2015 “Fitting Linear Mixed-Effects Models Using Lme4.” Journal of Statistical Software 67 (1): 1–48. CrossrefGoogle Scholar). Goodness-of-fit was assessed using marginal and conditional R2. Marginal R2 (R2m) excludes the random effects structure and conditional R2 (R2c) includes them. There was a main effect of linguistic variety (χ 2(2) = 23.68; p < 0.001), suggesting that there is indeed a difference among the three varieties regarding conducivity. The data was best fit when including the random effects structure (R 2c = 0.20). Delving further into the output of this model, we find that negative polarity tag questions in border Portuguese and border Spanish are more likely to be non-conducive than in Montevideo Spanish. The log odds of a non-conducive tag question were 1.57 +/− 0.33 standard errors (z = 4.73, p < 0.001) higher in border Portuguese than in Montevideo Spanish (a difference of approximately 34%). The same was true for border Portuguese (β = 0.99, SE = 0.33, z = 2.96, p < 0.004), which had roughly 75% non-conducive tag questions. In addition, negative polarity tag questions in border Portuguese are more likely to be non-conducive than border Spanish. Our model found a difference of −0.58 +/− 0.19 standard errors (z = −3.09, p < 0.003) in the log odds of using a non-conducive tag question (a decrease from approximately 82% to 75%). As shown in Figure 7, the percent of non-conducive negative polarity tag questions increases across varieties. While Montevideo Spanish shows the least tendency to use negative polarity tag questions with non-conducive functions, followed by Rivera Spanish, Rivera Portuguese shows the highest use of tags with non-conducive functions.

Figure 7.Percent non-conducivity of tag questions in Montevideo Spanish, border Spanish, and border Portuguese
Figure 7.
Montevideo Spanish Border Spanish Border Portuguese
  conducive     >         > non-conducive

Since non-conducivity is seen as a later stage in the development of negative polarity tag questions (Freitag 2008Freitag, Raquel Meister Ko 2008 “Marcadores Discursivos Interacionais na Fala de Itabaiana/SE.” Revista do Gelne, 10 (1): 21–32.Google Scholar; Pichler 2013 2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar), it is possible that these three linguistic varieties may be at three different stages in the trajectory of negative polarity tags, from least non-conducivity in monolingual Spanish to most non-conducivity in border Portuguese on the other extreme, while border Spanish is placed on the middle stage of this continuum. These different rates represent synchronic snapshots of the tendency seen in the use of Brazilian Portuguese (Freitag 2008Freitag, Raquel Meister Ko 2008 “Marcadores Discursivos Interacionais na Fala de Itabaiana/SE.” Revista do Gelne, 10 (1): 21–32.Google Scholar) and English innit (Pichler 2016 2016 “Uncovering Discourse-Pragmatic Innovations: Innit in Multicultural London English.” In Discourse-Pragmatic Variation and Change in English: New Methods and Insights, ed by Heike Pichler, 59–84. Cambridge, UK: Cambridge University Press. CrossrefGoogle Scholar) towards non-conducive uses of negative polarity tag questions.

The fact that Spanish-Portuguese bilinguals, compared with Spanish monolinguals, favor non-conducive functions over conducive functions, demonstrates that although border Spanish has not converged with Portuguese, it clearly shows a different behavior from its monolingual counterpart. This provides evidence for a certain permeability of discourse systems and the pragmatic borrowing of function (Andersen 2014 2014 “Pragmatic Borrowing.” Journal of Pragmatics 67: 17–33. CrossrefGoogle Scholar). The transfer of function without form has also been documented by several studies of equivalents of like in English in other languages (Hlavac 2006Hlavac, Jim 2006 “Bilingual Discourse Markers: Evidence from Croatian-English Code-switching.” Journal of Pragmatics 38 (11): 1870–1900. CrossrefGoogle Scholar; Kern 2014Kern, Joseph 2014 “Como in commute: The travels of a Discourse Marker across Languages.” Studies in Hispanic and Lusophone Linguistics 7 (2): 275–298. CrossrefGoogle Scholar; Maschler 2009Maschler, Yael 2009Metalanguage in Interaction: Hebrew Discourse Markers. Vol. 181. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar; Sankoff et al. 1997Sankoff, Gillian, Pierrette Thibault, Naomi Nagy, Hélène Blondeau, Marie-Odile Fonollosa, and Lucie Gagnon 1997 “Variation in the Use of Discourse Markers in a Language Contact Situation.” Language Variation and Change 9 (2): 191–217. CrossrefGoogle Scholar) and Zavala’s (2001)Zavala, Virginia 2001 “Borrowing Evidential Functions from Quechua: The Role of Pues as a Discourse Marker in Andean Spanish.” Journal of Pragmatics 33 (7): 999–1023. CrossrefGoogle Scholar study of the transfer of the discourse functions of the particle -mi in Quechua to pues in Spanish among Spanish-Quechua bilinguals in Peru. Although these results attest to a contact-induced tendency, they do not provide evidence for a generalized permeability of discourse systems, as claimed by Matras (1998Matras, Yaron 1998 “Utterance Modifiers and Universals of Grammatical Borrowing.” Linguistics 36 (2): 281–332. CrossrefGoogle Scholar, 2000 2000 “Fusion and the Cognitive Basis for Bilingual Discourse Markers.” International Journal of Bilingualism 4 (4): 505–528. CrossrefGoogle Scholar).

6.Conclusion

Both similarities and differences in the use of negative tags between these languages in contact were found among these bilingual participants. In terms of forms, while the overwhelming presence of ¿no? in bilingual Spanish presents a strong continuity with the monolingual dialect spoken in Montevideo, the infrequent but generalized use of ¿no es?, a calque from Portuguese, diverges from other Spanish monolingual norms. We further analyzed the use of negative tags in the Portuguese spoken by bilinguals. Although the generalized presence of negative tag ? adds evidence to previous claims that Uruguayan Portuguese and Brazilian Portuguese are closely related (Carvalho 2016 2016 “The Analysis of Languages in Contact: A Case Study through a Variationist Lens. Cadernos de Estudos Linguísticos 58 (3): 401–424. CrossrefGoogle Scholar), the presence of ¿no? as a direct borrowing from Spanish once again shows a tendency for minority languages to incorporate discourse-pragmatic features from the majority language. Although the borrowing from Spanish ¿no? was common in Uruguayan Portuguese and no es?, a calque from Portuguese was common in border Spanish, our analysis of both varieties does not provide evidence for the long-term replacement of an entire system found in Matras’ work (2000 2000 “Fusion and the Cognitive Basis for Bilingual Discourse Markers.” International Journal of Bilingualism 4 (4): 505–528. CrossrefGoogle Scholar, 514). A comparison of the functions of negative polarity tag questions between the Spanish spoken in Montevideo among monolinguals and in Rivera among Spanish-Portuguese bilingual showed a significant difference since bilinguals tended to use negative tags to fulfill non-conducive functions more often than monolinguals, a behavior hypothesized to have been influenced by contact with Portuguese. However, the significant difference found between the Spanish and Portuguese dialects spoken in Rivera shows that even though the tendency to use non-conducive negative polarity tag questions in border Portuguese may have influenced border Spanish (thus diverging it from its monolingual counterpart), these border varieties have not converged into a single system because of the significant difference found between them. Overall, these results suggest that even when cognate languages are in contact, discourse-pragmatic systems do not necessarily converge.

Studies of language contact outside of the field of language variation and change tend to exaggerate the effect of contact on dialects spoken by bilinguals, leading to the perception that contact-induced language changes are inevitable. In fact, Poplack, Zentz and Dion (2012Poplack, Shana, Lauren Zentz; and Nathalie Dion 2012Phrase-final Prepositions in Quebec French: An Empirical Study of Contact, Code-switching and Resistance to Convergence. Bilingualism: Language and Cognition 15(02): 203–225. CrossrefGoogle Scholar, 247) point out that “convergence is such a foregone conclusion of language contact that its very absence requires explanation”. As explained in our description of Rivera, speakers of Uruguayan Portuguese are constantly interacting with monolinguals of both Brazilian Portuguese and Southern Uruguayan Spanish. Carvalho (2014) 2014 “Sociolinguistic Continuities in Language Contact Situations: The Case of Portuguese in Contact with Spanish along the Uruguayan-Brazilian Border.” In Portuguese/Spanish Interfaces, ed. by Patrícia Amaral and Ana M. Carvalho, 263–294. Philadelphia: John Benjamins.Google Scholar suggests that urban border communities in Uruguay underwent rapid urbanization in the last part of the last century, which, in addition to the massive presence of monolingual urban varieties spoken in the media, has provided bilinguals with monolingual models, and more importantly, with a set of urban values that foments attitudes and ideologies that favor the diglossic separation of languages. In more general terms, the current analysis provides evidence that when bilingual dialects are compared with an appropriate reference following a variationist approach, preconceived assumptions of hybridization may be dismissed (Poplack and Levey 2010Poplack, Shana, and Stephen Levey 2010 “Contact-induced Grammatical Change: A Cautionary Tale.” In Language and Space: An International Handbook of Linguistic Variation 1, ed. by Jürgen Erich Schmidt and Peter Auer, 391–419. New York, NY: Walter de Gruyter.Google Scholar, 401).

In the future, it will be important to explore possible relationships between forms of negative polarity tag questions and their discourse functions, in line with Pichler’s (2010)Pichler, Heike 2010 “Methods in Discourse Variation Analysis: Reflections on the Way Forward 1.” Journal of Sociolinguistics 14 (5): 581–608. CrossrefGoogle Scholar suggestion to “include function as a parameter in the quantitative analysis and investigate whether functional variation and change impact on the distribution of variants” (590). The expansion of both corpora could also add apparent time data to detect a possible change in progress towards a more non-conducive use of negative tags in both Spanish and Portuguese spoken in Rivera. Following Pichler (2016) 2016 “Uncovering Discourse-Pragmatic Innovations: Innit in Multicultural London English.” In Discourse-Pragmatic Variation and Change in English: New Methods and Insights, ed by Heike Pichler, 59–84. Cambridge, UK: Cambridge University Press. CrossrefGoogle Scholar, it will also be important to determine to what extent non-conducive functions are found in non-canonical positions outside of the right periphery in order to assess to what extent the reduced form né? in Portuguese mirrors the path found for innit? in English. For now, by delimiting the variable according to its structural similarity, it was possible to (1) see that only a few forms are somewhat permeable, and to (2) capture an array of functions that are fulfilled by these forms, which revealed synchronic variation that is consistent with different diachronic stages toward the negative tag development across communities. We hope to have shed light on the subtleties of assessing the permeability of discourse-pragmatic features in language contact situations, and show that the assumed permeability of discourse-pragmatic features between cognate systems in intense, prolonged contact has not been attested in function or form.

Acknowledgements

The authors would like to express their gratitude to Joseph Casillas, for his generous and expert assistance with the statistical analysis, and to Adriana Picoral, for her invaluable help with coding, quantification, and formatting.

Note

1.Since no corpus of Sourthern Brazilian Portuguese is available, it is not possible to extend this comparison to include a monolingual variety of Portuguese.

References

Aaron, Jessi
2004 ““So Respetamos Un Tradicion Del Uno Al Otro.” So and Entonces in New Mexican Bilingual Discourse.” Spanish in Context 1 (2): 161–79. CrossrefGoogle Scholar
Andersen, Gisle
2001 “Pragmatic Markers and Sociolinguistic Variation a Relevance-theoretic Approach to the Language of Adolescents.” Pragmatics & Beyond 84. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing.Google Scholar
2014 “Pragmatic Borrowing.” Journal of Pragmatics 67: 17–33. CrossrefGoogle Scholar
Bates, Douglas, Martin Mächler, Ben Bolker, and Steve Walker
2015 “Fitting Linear Mixed-Effects Models Using Lme4.” Journal of Statistical Software 67 (1): 1–48. CrossrefGoogle Scholar
Behares, Luiz Ernesto
1984 “Diglosia en la Sociedad Escolar de la Frontera Uruguaya con Brasil: Matriz Social del Bilinguismo.” Cadernos de Estudos Linguísticos 6: 229–235.Google Scholar
Brody, Jill
1987 “Particles Borrowed From Spanish As Discourse Markers in Mayan Languages.” Anthropological Linguistics 29 (4): 507–521.Google Scholar
Carvalho, Ana Maria
2003 “The Sociolinguistic Distribution of (lh) in Uruguayan Portuguese: A Case of Dialectal Diffusion.” In Linguistic Theory and Language Development in Hispanic languages: Papers from the 5th Hispanic Linguistics Symposium and the 4th Conference on the Acquisition of Spanish and Portuguese, ed. by Silvina Montrul & Franscisco Ordóñez, 30–43. Somerville, MA: Cascadilla Press.Google Scholar
2004 “I Speak like the Guys on TV: Palatalization and the Urbanization of Uruguayan Portuguese.” Language Variation and Change 16 (2): 127–151. CrossrefGoogle Scholar
2006 “Spanish (s) Aspiration as a Prestige Marker on the Uruguayan-Brazilian Border.” Spanish in Context 3 (1): 85–114. CrossrefGoogle Scholar
2010 “¿Eres de la frontera o sos de la capital? Variation and Alternation of Second-person Verbal Forms in Uruguayan Border Spanish.” Southwest Journal of Linguistics 29 (1): 1–24.Google Scholar
2014 “Sociolinguistic Continuities in Language Contact Situations: The Case of Portuguese in Contact with Spanish along the Uruguayan-Brazilian Border.” In Portuguese/Spanish Interfaces, ed. by Patrícia Amaral and Ana M. Carvalho, 263–294. Philadelphia: John Benjamins.Google Scholar
2016 “The Analysis of Languages in Contact: A Case Study through a Variationist Lens. Cadernos de Estudos Linguísticos 58 (3): 401–424. CrossrefGoogle Scholar
Carvalho, Ana Maria and Bessett, Ryan
2015 “Subject Pronoun Expression among Spanish-Portuguese Bilinguals.” In Subject Personal Pronouns in Spanish: a Cross-dialectal Perspective, ed. by Ana Maria Carvalho, Rafael Orozco and Naomi Lapidus Shin, 275–315. Washington D.C.: Georgetown University Press.Google Scholar
Castañeda-Molla, Rosa Maria
. Linguistic Variation in a Border Town: Palatalization of Dental Stops and Vowel Nasalization in Rivera. PhD diss., University of Florida 2011.Google Scholar
Córdoba, Alexander Severo
2017 “O Português Uruguai Falado em Tranqueras-Uruguai: O Fenômeno da Elevação das Vogais Postônicas Finais.” Web Revista SOCIODIALETO 7 (20): 389–427.Google Scholar
Dajko, Nathalie, and Katie Carmichael
2014 “But Qui C’est La Différence ? Discourse Markers in Louisiana French: The Case of but vs. mais.” Language in Society 43 (2): 159–83. CrossrefGoogle Scholar
de Rooij, Vincent
2000 “French Discourse Markers in Shaba Swahili Conversations.” International Journal of Bilingualism 4 (4): 447–67. CrossrefGoogle Scholar
Domínguez Mujica, Carmen Luisa
2005 “Marcadores de (In)conclusión en el Español Hablado en Mérida-Venezuela. Boletín De Lingüística, 23: 3–22.Google Scholar
Elizaincín, Adolfo
1995 “Personal Pronouns for Inanimate Entities in Uruguayan Spanish in Contact with Portuguese.” In Spanish in Four Continents: Studies in Language Contact and Bilingualism, ed. by Carmen Silva-Corvalán, 117–131. Washington, D.C.: Georgetown University Press.Google Scholar
Elizaincín, Adolfo, Luis Ernesto Behares, and Graciela Barrios
1987Nos falemo brasilero: dialectos portugueses en Uruguay. Montevideo, Uruguay: Editorial Amesur.Google Scholar
Ferreira, Madalena
1981 “Tag Questions in Portuguese: Grammar and Intonation.” Phonetica 38 (5): 341–352. CrossrefGoogle Scholar
Flores-Ferrán, Nydia
2014 “So Pues Entonces: An Examination of Bilingual Discourse Markers in Spanish Oral Narratives of Personal Experience of New York City-born Puerto Ricans.” Sociolinguistic Studies 8 (1): 57–79. CrossrefGoogle Scholar
Freitag, Raquel Meister Ko
2008 “Marcadores Discursivos Interacionais na Fala de Itabaiana/SE.” Revista do Gelne, 10 (1): 21–32.Google Scholar
Fuller, Janet M.
2001 “The Principle of Pragmatic Detachability in Borrowing: English-origin Discourse Markers in Pennsylvania German.” Linguistics 39 (2): 351–370. ISSU 372. CrossrefGoogle Scholar
Gao, Hua
2013 “Tag-Questions in Mandarin Chinese.” In Grammar in Cross-Linguistic Perspective The Syntax, Semantics, and Pragmatics of Japanese and Chinese ed. by Teruhiro Ishiguro and Kang Kwong Luke, 43–64. Bern, Switzerland: Peter Lang AG.Google Scholar
García Vizcaíno, María José
2005 “El uso de los apéndices modalizadores¿ no? y¿ eh? en español peninsular.” In Selected Proceedings of the Second Workshop on Spanish Sociolinguistics ed. by Lotfi Sayahi and Maurice Westmoreland, 89–101. Somerville, MA: Cascadilla Proceedings Project.Google Scholar
Gómez-González, María de los Ángeles
2014 “Canonical Tag Questions in English, Spanish and Portuguese: A Discourse-functional Study.” Languages in Contrast 14 (1): 93–126. CrossrefGoogle Scholar
Gorski, Edair Maria and Raquel Meister Ko Freitag
2006 “Marcação e Comportamento Sociolinguístico de Marcadores Discursivos Interacionais de Base Verbal na Fala de Florianópolis.” In Variação e mudança linguística na Região Sul, ed. by Paulino Vandresen, 29–48. Pelotas, Brazil: EDUCAT.Google Scholar
Hlavac, Jim
2006 “Bilingual Discourse Markers: Evidence from Croatian-English Code-switching.” Journal of Pragmatics 38 (11): 1870–1900. CrossrefGoogle Scholar
Kern, Joseph
2014 “Como in commute: The travels of a Discourse Marker across Languages.” Studies in Hispanic and Lusophone Linguistics 7 (2): 275–298. CrossrefGoogle Scholar
Kimps, Ditte, Kristin Davidse, and Bert Cornillie
2014 “The Speech Functions of Tag Questions and their Properties. A Comparison of their Distribution in COLT and LLC.” Corpus Interrogation and Grammatical Patterns, 321–50. Amsterdam, Netherlands: John Benjamins. CrossrefGoogle Scholar
Leite, Ligia Chiappini Moraes, and Maria Helena Martins
2006Cone Sul: Fluxos, Representações e Percepções Vol. 38. Editora Hucitec.Google Scholar
Lim, Lisa, and Umberto Ansaldo
2016Languages in Contact. Cambridge, UK: Cambridge University Press. CrossrefGoogle Scholar
Lipski, John M.
2006 “Too Close for Comfort? The Genesis of “Portuñol/Portunhol”.” In Selected proceedings of the 8th Hispanic Linguistics Symposium 1: 22–54. Somerville: Cascadilla Proceedings Project.Google Scholar
2009 “Searching for the Origins of Uruguayan Fronterizo Dialects: Radical Code-mixing as “Fluent Dysfluency”. Journal of Portuguese Linguistics 8 (1): 3–44. CrossrefGoogle Scholar
Marín, Francisco Marco
2001 “De lenguas y fronteras. El Spanglish y el portuñol.” Nueva Revista de Política, Cultura y Arte, 74: 72–79.Google Scholar
Maschler, Yael
2009Metalanguage in Interaction: Hebrew Discourse Markers. Vol. 181. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar
Matras, Yaron
1998 “Utterance Modifiers and Universals of Grammatical Borrowing.” Linguistics 36 (2): 281–332. CrossrefGoogle Scholar
2000 “Fusion and the Cognitive Basis for Bilingual Discourse Markers.” International Journal of Bilingualism 4 (4): 505–528. CrossrefGoogle Scholar
Meirelles, Virginia Garrido
2009 “O Português da Fronteira Uruguai-Brasil.” In Português em Contato, ed. by Ana M. Carvalho, 257–76. Madrid: Iberoamericana. CrossrefGoogle Scholar
Mendes, Ronald Beline
2013Projeto SP2010: Amostra da Fala Paulistana. http://​projetosp2010​.fflch​.usp​.br
Mougeon, Raymond, and Edouard Beniak
1991Linguistic Consequences of Language Contact and Restriction: The Case of French in Ontario, Canada. Oxford: Clarendon.Google Scholar
Muysken, Pieter, Carmen Pena Díaz, and Pieter Cornelis Muysken
2000Bilingual Speech: A typology of Code-mixing. Vol. 11. Cambridge, UK: Cambridge University Press.Google Scholar
Myers-Scotton, Carol, and Janice L. Jake
1995 “Matching Lemmas in a Bilingual Language Competence and Production Model: Evidence from Intrasentential Code Switching.” Linguistics 33 (5): 981–1024. CrossrefGoogle Scholar
Ocampo, Francisco
2013 “La Discursivización de Viste/Vio en Rioplatense.” In Perspectivas teóricas y experimentales sobre el español de la Argentina, ed. by Laura Colantoni and Celeste Rodríguez Louro, 487–505. Madrid: Iberoamericana. CrossrefGoogle Scholar
Pacheco, Cintia da Silva
2017 “Identidade sociolinguística na fronteira de Aceguá (Brasil-Uruguai).” Revista de Estudos da Linguagem 25 (1): 276–304. CrossrefGoogle Scholar
Peterson, Elizabeth, and Johanna Vaattovaara
2014 “Kiitos and pliis: The Relationship of Native and Borrowed Politeness Markers in Finnish.” Journal of Politeness Research 10 (2): 247–269. CrossrefGoogle Scholar
Pichler, Heike
2010 “Methods in Discourse Variation Analysis: Reflections on the Way Forward 1.” Journal of Sociolinguistics 14 (5): 581–608. CrossrefGoogle Scholar
2013The Structure of Discourse-pragmatic Variation. Vol. 13. Amsterdam, Netherlands; Philadelphia, PA: John Benjamins Publishing. CrossrefGoogle Scholar
2016 “Uncovering Discourse-Pragmatic Innovations: Innit in Multicultural London English.” In Discourse-Pragmatic Variation and Change in English: New Methods and Insights, ed by Heike Pichler, 59–84. Cambridge, UK: Cambridge University Press. CrossrefGoogle Scholar
Poplack, Shana, Lauren Zentz; and Nathalie Dion
2012Phrase-final Prepositions in Quebec French: An Empirical Study of Contact, Code-switching and Resistance to Convergence. Bilingualism: Language and Cognition 15(02): 203–225. CrossrefGoogle Scholar
Poplack, Shana, and Stephen Levey
2010 “Contact-induced Grammatical Change: A Cautionary Tale.” In Language and Space: An International Handbook of Linguistic Variation 1, ed. by Jürgen Erich Schmidt and Peter Auer, 391–419. New York, NY: Walter de Gruyter.Google Scholar
PRESEEA
. Corpus del Proyecto para el estudio sociolingüístico del español de España y de América. Alcalá de Henares: Universidad de Alcalá. http://​preseea​.linguas​.net 2014.
R Core Team
2013R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, http://​www​.R​-project​.org
Rodríguez-Muñoz, Francisco J.
2009 “Estudio sobre las funciones pragmadiscursivas de¿ no? y¿ eh? en el español hablado.” Revista de Lingüística Teórica y Aplicada 47 (1): 83–101.Google Scholar
Salmons, Joe
1990 “Bilingual Discourse Marking: Code Switching, Borrowing, and Convergence in Some German-American Dialects.” Linguistics 28 (3): 453–480. CrossrefGoogle Scholar
Sankoff, Gillian, Pierrette Thibault, Naomi Nagy, Hélène Blondeau, Marie-Odile Fonollosa, and Lucie Gagnon
1997 “Variation in the Use of Discourse Markers in a Language Contact Situation.” Language Variation and Change 9 (2): 191–217. CrossrefGoogle Scholar
Silva, Giselle M. de Oliveira, and Alzira Tavares de Macedo
1992 “Discourse Markers in the Spoken Portuguese of Rio De Janeiro.” Language Variation and Change 4 (2): 235–49. CrossrefGoogle Scholar
Sturza, Elaine
2004 “Fronteiras e Práticas Linguísticas: Um Olhar sobre o Portunhol.” Revista Internacional de Linguística Iberoamericana 3: 151–162.Google Scholar
Tagliamonte, Sali A.
2013 “Comparative Sociolinguistics.” In The Handbook of Language Variation and Change Vol. 129, ed. by Jack K. Chambers and Natalie Schilling: 128–156. New York, NY: John Wiley & Sons. CrossrefGoogle Scholar
Tomaselli, Maria Vittoria, and Albert Gatt
2015 “Italian Tag Questions and their Conversational Functions.” Journal of Pragmatics 84: 54–82. CrossrefGoogle Scholar
Torres, Lourdes
2002 “Bilingual Discourse Markers in Puerto Rican Spanish.” Language in Society 31 (1): 65–83. CrossrefGoogle Scholar
Torres, Lourdes, and Kim Potowski
2008 “A Comparative Study of Bilingual Discourse Markers in Chicago Mexican, Puerto Rican, and MexiRican Spanish.” International Journal of Bilingualism 12 (4): 263–279. CrossrefGoogle Scholar
Vázquez Carranza, Ariel
2017 “Some Uses of ‘no’ in Spanish Talk-in-Interactions.” International Review of Pragmatics 9, (2): 224–247. CrossrefGoogle Scholar
Zavala, Virginia
2001 “Borrowing Evidential Functions from Quechua: The Role of Pues as a Discourse Marker in Andean Spanish.” Journal of Pragmatics 33 (7): 999–1023. CrossrefGoogle Scholar
Waltermire, Mark
2006Social and Linguistic Correlates of Spanish-Portuguese Bilingualism on the Uruguayan-Brazilian border. New Mexico: ProQuest.Google Scholar
2011 “Frequency Effects on the Morphological Conditioning of Syllable-Final /s/ Reduction in Border Uruguayan Spanish.” Journal of Language Contact 4 (1): 26–55. CrossrefGoogle Scholar

Address for correspondence

Ana M. Carvalho

University of Arizona

Modern Languages 545

Tucson, AZ 85721

United States

anac@email.arizona.edu

Biographical notes

Ana Carvalho is a Professor of Hispanic Linguistics at the University of Arizona. She works with languages in contact from a sociolinguistic perspective, and is especially interested in patterns that evolve in situations of contact between cognate languages, such as Spanish and Portuguese.

Joseph Kern is an Assistant Professor of Spanish at The University of Virginia’s College at Wise. His research explores discourse-pragmatic variation in Spanish, especially in varieties of Spanish in contact. He has analyzed several discourse-pragmatic variables in Spanish including quotatives, discourse markers, general extenders, and tag questions.