Langues et civilisations
à tradition orale

          Le CNRS           Accueil SHS          Autres sites CNRS          
cnrs paris3 inalco paris3 paris3
Intranet Lacito English

  Accueil > Colloques > 13th International Conference on Romani Linguistics (ICRL13))





13ème Conférence Internationale de Linguistique Romani (ICRL13)
13-14 septembre 2018



Résumés en anglais (pdf ici)



Conceptual transfer in copula choice: More evidence from the Mexican Romani-Spanish bilinguals toup

Evangelia Adamou
CNRS (French National Centre for Scientific Research)

Stefano de Pascale
KU Leuven

Yekaterina Garcia-Markina
University of Tours

Cristian Padure

There is ongoing discussion as to whether bilinguals generalize the Spanish copula estar "to be" as a result of general cognitive processes and the extent to which conceptual transfer from the contact language plays a role (e.g., Silva-Corvalán 1986, Geeslin & Guijarro-Fuentes 2008). This study aims to tease apart these factors by investigating copula choice in Romani and in Spanish spoken in Mexico.

To allow for comparability with previous studies, we used the contextualized copula choice task designed by Geeslin & Guijarro-Fuentes (2008), comprising of 28 sentences. 60 Mexican Romani- Spanish bilinguals responded to the questionnaire in both Spanish and Romani, and 62 Mexican Spanish monolinguals responded in Spanish. Generalized linear mixed effects models were constructed to analyze the results.

Analysis of the results reveals greater extension of estar among the bilinguals for individual-level predicates as well as for traits not susceptible to change. Comparison of the responses of the bilinguals (in Romani and in Spanish) and of the Spanish monolinguals indicates that Romani could be reinforcing the generalization of estar in the Spanish responses of the bilinguals.

In sum, this is the first study to examine Spanish copula choice in the bilingual mode by also looking at the results in the contact language. The Romani-Spanish data support the hypothesis that bilinguals generalize estar faster than monolinguals. In addition, these results suggest that the equivalence between the two languages is the driving effect for the generalization of estar among bilinguals (also see Adamou 2013).

-- Adamou, E. 2013. Replicating Spanish estar in Mexican Romani. Linguistics 51(6): 1075–1105.
-- Geeslin, K. & P. Guijarro-Fuentes. 2008. Variation in contemporary Spanish: Linguistic predictors of estar in four cases of language contact. Bilingualism: Language and Cognition 11(3): 365–380.
-- Silva-Corvalán, C. 1986. Bilingualism and language change: The extension of estar in Los Angeles Spanish. Language 62: 587–608.


On the Origin and development of Spanish and Catalan Para-Romani varieties toup

Ignasi-Xavier Adiego
Universitat de Barcelona

In this paper a comprehensive overview is offered of the currently available information on Spanish Romani and Catalan Romani regarding the question of their origin and development. After the important works of Bakker (1994), (1998), new materials, revised old documentation and field work make necessary to proceed to update and re-evaluate all this information. Catalan Romani and Spanish Romani show similiarities in their development, but also significant differences, particularly in connection with the controversy on gradual evolution vs. sudden apparition of Para-Romani dialects.

-- P. Bakker. 1994. «Notes on the genesis of Caló and other Iberian Para-Romani varieties», in Y. Matras, ed., Romani in Contact, Amsterdam-Philadelphia,125-150.
-- P. Bakker. 1998. «Para-Romani languages versus secret languages: Differences in origin, structure and use», In Y. Matras, ed. The Romani Element in Non-Standard Speech, Wiesbaden, 69-96.


The Sinti dialect in Hungary toup

Márton András Baló
Research Institute for Linguistics, Hungarian Academy of Sciences

The Sinti, albeit constituting a small proportion of the Romani people in Hungary, stand out as a peculiar subgroup among the more numerous and more dominant Central and Vlax Romani speakers in Hungary. Compared to other Romani groups, the Sinti are relative latecomers to Hungary: they first arrived in Tolna county around the turn of the 19th and 20th centuries from Western Europe, working as itinerant craftsmen, looking for new places where their skills were in demand. Later, other groups fleeing Nazi persecution headed for Hungary again. The majority of them settled after the Second World War and now they live scattered across the country in such diverse places as Pomáz, Ozora and its area (Tamási, Gyönk), Paks, Gyöngyös and the Balaton Uplands. Their precise number is difficult to tell, but some estimates put their number at a few hundred, while others presume that there are a few thousand of them. Their dialect seems to be on the brink of extinction, as the younger generations do not really speak it any more; a rapid language shift has taken place. So far, only one attempt has been made at describing the Sinti dialect spoken in Hungary, by György Mészáros in 1980. In our paper, we would like to present the linguistic differences and similarities between the state of the Sinti dialect as described in 1980 and the state as it is spoken today, while also comparing its phonology, morphology and syntax to other Sinti varieties spoken in other parts of Europe. To that end, we are planning to use newly collected data. New data have been collected already with the help of the Linguistic Questionnaire for the Documentation of Central European Romani (LQCR), as well as the questionnaire used to document morphological variation in Vlax Romani in Hungary (Research Institute for Linguistics, Hungarian Academy of Sciences), and more data are planned to be collected with the help of the the Romani Dialectological Questionnaire (RMS) and the revised version of the RIL HAS questionnaire, which are then going to be painstakingly analysed.

The significance of the research underway is immense: while documenting a Romani variety that is virtually disappearing, we can also gain insight into the extent of language contact on the one hand, comparing it to the contact phenomena seen in other Romani varieties spoken in Hungary, and look at the possible differences in language change across Sinti dialects spoken in different geographical areas of Europe.


Semantic change in Romani terms denoting human beings toup

Zuzana Bodnárová
University of Graz

The division of the world to the Roma on the one hand and all the others on the other, thus we as opposed to they, represents a core cultural concept of the Roma. This concept is reflected not only in Romani but also often in the various contact language(s) spoken by the Roma. For instance, non-Roma are referred to by the Prekmurje Roma as Pavri in the Prekmurje dialect of Slovenian as opposed to Romi 'Roma', while the corresponding Romani terms are Gourgja 'non-Roma' and Rouma 'Roma'.

Romani terms which have originally encoded such an ethnic division are terms denoting the human beings 'woman', 'man', 'girl' and 'boy', including the kinship terms 'wife', 'husband', 'daughter' and 'son' (see Beníšek 2006). In most Romani dialects spoken today, however, a part of these terms have undergone some type of semantic change, e.g. specialisation and generalisation of meaning or antiphrasis (for the types of change see e.g. Blank 1999).

Based on data from various Romani dialects (online and printed dictionaries, own research), I will present by means of semantic maps the semantic domains of human beings and kin terms and their lexicalisation in the frames of 'Romani ethnicity', 'non-Romani ethnicity' and 'ethnic neutral'. The maps will show lexemes that are used to cover these semantic domains and the patterns of semantic changes that took place in the individual Romani dialects (e.g. the loss of Romani-specific terms in Vend Romani). Finally, I will look into possible sociocultural motivations across the dialects for such changes as well as discuss some external factors accounting for it such as language contact, language endangerment, and geographical isolation.

-- Beníšek, Michael. 2006. Ke kořenům slova ROM. Romano džaniben, jevend 2006: 9–27.
-- Blank, Andreas. 1999. Why do new meanings occur? A cognitive typology of the motivations for lexical Semantic change. In: Blank, Andreas / Koch, Peter: Historical Semantics and Cognition, Berlin/New York: Mouton de Gruyter, 61–90.


Loan-verb adaptation in Gurbet Romani in Knjaževac (Eastern Serbia) toup

Svetlana Ćirković
Mirjana Mirić
Institute for Balkan Studies, Serbian Academy of Sciences and Arts, Belgrade (Serbia)

The aim of the study is to provide a model of Serbian loan-verb adaptation to the Gurbet Romani variety (Southern Vlax group of Romani dialects (Matras 2004)), spoken in the town of Knjaževac and its surrounding area (Eastern Serbia). According to 2011 Census, 673 out of 31,491 people are using Romani in this area, where extensive language contact between Serbian as a dominant language and different varieties of Romani is reported (Sikimić 2017, Ćirković & Mirić 2017).

Previous research on Romani loan-verbs (Boretzky 2003, Matras 2004, Leggio 2011, Leggio & Matras 2017) has shown that in the Vlax varieties borrowed verbal roots are adapted by using the marker -isar- for transitives and -isav- for intransitives, displayed as -sard- and -sajl- in the past tense. An optional reduction of markers to their vocalic component -i/-o was also reported (Leggio 2011, Leggio & Matras 2017).

The corpus for our research contains transcripts of conversations with 12 native speakers of the Gurbet variety recorded in Knjaževac and the area in July 2016. All speakers are bilingual in Romani and Serbian. The corpus contains ca. 45,000 word tokens. The open-ended interview was used as a method of data collection, with questions focusing mainly on traditional culture.

The corpus was manually lemmatized. If the basic form of the verb (3.sg.pres.) or the past tense (3.sg. past.) was not found in the corpus, 3 native speakers were additionally asked to reconstruct those forms. Two verb groups were excerpted: a) Romani verbs (N=151; frequency=652); b) Serbian loan- verbs, morphologically adapted to Romani (N=180; frequency=361). For the purpose of this study, only the loan-verbs are analyzed.

Loan-verbs are built from the Serbian verb root to which loan-verb adaptation markers are added. The analysis revealed the following tendencies:
     a) The prevalent reduction of the adaptation marker -sar- to -i/-o in the present tense and the retention of the marker -sard- in the past tense in 72.8% of verbs: transitives – 80.2%, intransitives – 11.4%, reflexive/reciprocal – 8.4%; e.g. cenil–cenisarda 'appreciate', čistol– čistosarda 'clean'.
     b) The retention of the marker -sar- in the present tense only in doublet forms of transitives (5%); e.g. pričol, pričosarel 'tell'.
     c) The reduction of the marker -sáv- to -i/-o in the present tense and the retention of the marker -salj- (<-sajl-) in the past tense in 7.8% of verbs: transitives – 42.85%, intransitives – 14.3%, reflexive/reciprocal – 42.85%; e.g. snalazil pe - snalazisaljo'manage'.
     d) The tendency of mainly reflexive/reciprocal verbs to form the past tense with the marker -salj-, as well as -sard- in analytic reflexive constructions with oblique pronouns or the reflexive pronoun pe(s) (14.4%); e.g. mučil pe - mučisaljo/ mučisarda pe 'suffer'. e) The redundancy of transitivity-intransitivity distinction, given the inconsistent use of the adaptation markers -sard- / -salj- regarding the transitivity (letil IN TRAN - letisarda 'fly', pokažil TRAN - pokažisaljo 'show', 'point'), and the overall tendency of speakers to use the marker -sard- in the past tense of intransitives and reflexives by analogy with transitives.

The results will be discussed in comparison to the previous findings on loan-verbs in Vlax Romani varieties, particular attention being paid to the changes in the adaptation markers system.


Iotation and its morphophonological effects in a Romani variety toup

Viktor Elšík
Department of Linguistics, Charles University

The paper will present an analysis of the so-called IOTATION and its morphophonological effects in a variety of South Central Romani spoken in a multiethnic village in south-western Slovakia. (The data have been acquired through the author's long-term linguistic fieldwork in the locality.) In many dialects of Romani, including the variety in question, a number of suffixes may be analyzed as beginning in an underlying palatal approximant /j/, whose elision triggers a morphophonological alternation of the preceding consonant, e.g. pheň-a [pheŋa] <phen-ja> [sister-NOM.PL] 'sisters' (← phen 'sister'). The Romani variety in question stands out in exhibiting a complex pattern of iotation morphophono-logy, with several different effects in differing structural contexts.

In this paper I will first present an overview of the grammatical distribution of the under-lying palatal approximant, viz. in noun inflection, perfective verb inflection, and verb derivation. Second, I will illustrate the different effects of iotation, viz. a) palatalization, e.g. angrusť-a ['aŋgrusca] <angrust-ja> [ring-NOM.PL] 'rings' (←angrust-i 'ring'); b) gemination, e.g. posiťť-a ['posic:a] <posiť-ja> [pocket-NOM.PL] 'pockets' (←posiť-i'pocket'); and c) palatalization and gemination, e.g. búťť-a ['bu:c:a] <bút-ja> [thing-NOM.PL] 'things, works' (←bút-i 'thing, work'). Third, I will analyze the phonological and grammatical conditioning of the different effects of iotation. And, finally, I will show that some of the iotation rules in this Romani variety have been copied from Hungarian (cf. e.g. Siptár & Törkenczy 2000) and briefly discuss the role of language contact in complexifying the morphophonology of a language.

NOM = nominative
PL = plural

-- Siptár, Péter & Törkenczy, Miklós. 2000. The phonology of Hungarian. Oxford: Oxford University Press.


The Problem of Long versus Short Present Tense forms in Balkan Romani toup

Victor A. Friedman
University of Chicago & La Trobe University

Matras (2002:117-118, 156 et passim) states that the difference between long versus short form present tense in -a (long) vs -Ø (short) in the Romani of the Balkans is that of indicative versus subjunctive, e.g. me kerava 'I do/am doing' vs me te kerav/me ka kerav 'let me do~that I do/I will do'. On the one hand, this is in contrast to the opposition long vs short as future vs present in many dialects outside the Balkans (Boretzky and Igla 2004:map 138), and, on the other hand, it is in contrast to Cech and Heinschink's (2002) observation that such a distinction does not exist in Prishtina Arli and related varieties, although they do not propose an alternative explanation. Boretzky (1993:177-178) also attests the same type of variation as in Prishtina Arli in the Burgudži (Bugurdži) dialect of Kosovo, and the same variation is evident in Cech, Heinschink and Halwachs (2009:168 et passim).

I have also observed this lack of distinction in Arli dialects in Macedonia. The lack of distinction in Prishtina Arli and Kosovo Burgudži has significant implications for language contact in Kosovo prior to World War Two, and overall the phenomenon seems to be related to contact with Balkan Slavic. Nonetheless, the fact remains that both long and short forms are well entrenched in all varieties of Romani spoken in the Balkans, and they often follow the expected rules. Nonetheless, if a straightforward indicative/subjunctive opposition does not explain all the usages in at least some dialects, then the possibility of some sort of pragmatic conditioning should be investigated. It is likely that focus, emphasis, and/or topicalization have roles to play. I do not yet have more than this tentative answer to the question posed, but I am in the process of conducting the analysis and hope to present my results at the conference.

I have both my own original field recordings collected in recent decades as well as published sources and am investigating them to provide at least a tentative answer to the question of what conditions the long/short present tense opposition in Balkan Romani dialects where indicative/ subjunctive is clearly not the conditioning factor. The paper will seek to explicate pragmatic conditioning factors other than the grammatical, since these latter clearly fail to explicate usage. The significance of the results will lie in opening up new possibilities for investigating the long/ short opposition in Romani dialects of the Balkans.

-- Boretzky, Norbert. 1993. Bugurdži: Deskriptiver und historischer Abriß eines Romani-dialekts. Wiesbaden: Harrassowitz.
-- Boretzky, Norbert & Birgit Igla. 2004. Komentierter Dialektatlas des Romani. Teil 1: Vergleich der Dialekte. Teil 2: Dialektkarten mit einer CD Rom. Wiesbaden: Harrassowitz.
-- Cech, Petra and Mozes F Heinschink. 2002. "The Arli dialect of Priština and other Arli varieties spoken in Serbia and Macedonia." Paper presented at the 6th International Conference on Romani Linguistics, Center for the Study of Modern European Languages and the University of Graz, Austria, 12-14 September 2002.
-- Cech, Petra, Moses F. Heinschink, Dieter W. Halwachs, eds. 2009. Momelja hem limonja: Arlijengere paramisja. Klagenfurt/Celovec: Drava Verlag.
-- Matras, Yaron. 2002. Romani: A linguistic introduction. Cambridge: Cambridge University Press.


Standardized Romani from Romania from the perspective of the native speakers toup

Adrian-Nicolae Furtuna
National Centre of Roma Culture

The subject of standardization of the Romani language is one that has generated polemics ever since its inception, either through the standardization strategy (eg, Hancock 1975) or the alphabet set by IRU 1990, insufficient for a faithful scoring of the phonetic particularities of each dialect (eg. Matras 2004).

Our research is focused on the Romani language in Romania and aims, on the one hand, to highlight the main linguistic and cultural elements that make the native speakers (not) to identify themselves with this standardized form of the Romani language and, on the other hand, analyze to what extent standardized language is found in community daily communication after more than 25 years of contact.

The methodology of data collection addresses a series of questions regarding the level of understanding and appropriation of the standard Romani language by native speakers of Romani varieties attested in Romania (eg, Elšík & Matras 2006: 55-64).

The research sample consists of 40 people, native speakers of Romani, aged 14 - 25, members of communities with tradition in teaching the Romani language in school and in the family.

-- Elšik, V., & Matras, Y. (2006). Markedness and Language Change: The Romani Sample. Walter de Gruyter.
-- Hancock Ian, 1975, Problems in the Creation of a Standard Dialect of Romanes, Social Science Research Council on Sociolinguistics Working Paper N0, 25.
-- Matras, Y. (2004). Romacilikanes—The Romani dialect of Parakalamos. Romani Studies, 14(1), 59-109.


Scenarios of complex linguistic variation and change among Arli speakers in Southern Sweden toup

Kimmo Granqvist
Södertörn University, Stockholm

Viveka Sajin
Södertörn University, Stockholm

Mirkka Salo
University of Helsinki

In our paper, we will report findings of a case study of scenarios of complex linguistic variation among a group of Arli speakers living in urban areas in Southern Sweden. These Roma migrated to Sweden in the early 1990s from Kosovo and Serbia. They speak Romani as their native language, they are fluent in Serbian, and possess varying insights in Swedish and other languages such as English. Some of them also master Albanian and Turkish.

A mixture of qualitative and quantitative methods will be used for data collection. We will collect linguistic data by relying on participant observation and recording10 semi-spontaneous autobiographical narratives, based on which we will create an annotated corpus. We will complement this data by carrying out qualitative interviews of the informants on language attitudes, their networks and linguistic history information.

We will discuss the communicative space of the Arli speakers in Southern Sweden, adopting a multidimensional and dynamic model (Krefeld 2010). We will examine the negotiation of prestige in multi-language and multi-dialect settings combining linguistic and social aspects. We will analyze our data from the perspective of language variation and change, dialect leveling, the convergence between diverse Romani varieties and the emergence of possible new varieties.

Our theoretical focus will be partly based on migration or migratory linguistics (Krefeld 2004). Wewill furthermore draw on urban sociolinguistics (Labov 1972; Trudgill 1974; Milroy 1987; Chambers & Trudgill 1998) and functional typology. We seek to contribute to migration linguistics and sociolinguistic theory. We wish also to contribute to the work on the Romani language in migrant settings (e.g. Petrović & Stefanović 2005; Matras 2015) and contact linguistics related to Romani.

-- Adamou E. 2010. Bilingual Speech and Language Ecology in Greek Thrace: Romani and Pomak in contact with Turkish. Language in Society 39(2), 147-171.
-- Chambers, J.K. & Trudgill, P. 1998. Dialectology. New York: Cambridge University Press.
-- Krefeld, Th. 2004. Einführung in die Migrationslinguistik: Van der Germania italiana in die Romania multipla. Tübingen: Narr.
-- Krefeld, Th. 2010. The consequences of migration and colonialism III: New minorities. In: P. Auer and J. E. Schmidt (eds.) Language and space: an international handbook of linguistic variation. New York: De Gruyter Mouton. 468-478.
-- Labov, W. 1972a. Language in the Inner City: Studies in the Black English Vernacular. Philadelphia: University of Pennsylvania Press.
-- Matras, Y. 2015. Transnational policy and 'authenticity' discourses on Romani language and identity. Language in Society 44, 295-312.
-- Milroy, L. 1987. Language and Social Networks. Oxford: Blackwell.
-- Petrović, J. & Stefanović, L. 2005. Sociolinguistic aspects of the language of Roma refugees from Kosovo - A comparative study. In: B. Schrammel, D.W. Halwachs & G. Ambrosch (eds.), General and Applied Romani Linguistics. Proceedings from the 6th International Conference on Romani Linguistics. Munich: Lincom Europa. 174-181.
-- Trudgill, P. 1974. The Social Differentiation of English in Norwich. Cambridge: Cambridge University Press.


Romani: Labels - Status – Situation toup

Dieter Halwachs
University of Graz

Triggered by a rather positive and open-minded political climate, as well as a liberal attitude towards plurality and minorities during the last decades of the 20th century, marginalised and discriminated groups like the Roms were encouraged and enabled to start a process of self- organisation, with the further aim of emancipation. Consequently, Romani history, culture, and language attracted the attention of both academia and politics and became, more or less, part of the public consciousness. For the first time in European history, the Romani language was, at least to a certain extent, recognised and perceived as an integral part of a common European linguistic heritage. Although characterised by a high degree of variation and heterogeneity, without any written tradition and public usage, Romani was judged on the basis of the European common-sense definition of a language, which presupposes a standard as well as a literate style that functions in public formal domains. Since such a linguistic state has always been a desideratum on the part of the Romani movement, its representatives not only claim the need for a Romani standard, but sometimes even insinuate its existence. Against the background of this contradictory situation, Romani became a flamboyant phenomenon, not only among the languages of Europe but worldwide. In both academic and popular contexts, the labels used to characterise the Romani language are manifold, ranging from indigenous language to migrant language, from community language to national language, from local language to international language, etc. Thus, Romani is often perceived and treated as something extraordinarily special among languages.

Starting from an outline of its sociolinguistic situation and a summary of its complex socio-political status, the paper reviews the content and validity of the manifold labels Romani is tagged with and, furthermore, discusses its reputed and reported uniqueness among the languages of the world.


Innovating the past: the remoteness marker -esta in Shinto Rosengro toup

Giulia Meli
Università degli Studi di Milano

The discovery of Shinto Rosengro brings about new data to the panorama of old settlement Romani varieties in Italy. Discovered by Leonardo Piasere in the late 90s' and partly published in 2001, the manuscripts witnessing this dialect were arranged from 1892 to 1911 by Sigismondo Caccini, a man who had a strict contact with the community of Shinte Rosengre and who took part to its short-range movements through Central Italy. The texts and grammatical descriptions by Caccini show a Sinti variety carrying some peculiar innovations, which distinguish it from the other old settlement varieties on the Italian ground (both Sinti and Abruzzian Romani).

One of those innovations is the remoteness marker -esta, used to form the imperfect and pluperfect. When it is added to the present, it forms the imperfect, with the exception of the third person singular, which is built through the addition of the remoteness marker -esta to the bare lexical morpheme of the verb, es. kil- én-esta 'dance-3PL-REM''they danced', kil-ésta 'dance-3S.REM''he danced'. As regularly happens in many other varieties, the remoteness morpheme -esta is added to the forms of the preterite to build the pluperfect es. kamj-ón-esta 'want.PF-1S-REM''I had wanted' vs. kami-ón 'want.PF-1P''I wanted'.

This remoteness marker seems unique among the varieties known up to now. The first part -es-can be reconnected to the more spread remoteness markers attested in romani, -as/-a/-e/-s/-ys/-ahi (cfr. Matras 2002: 170; Boretzky & Igla 2004) originated from a copula form (cfr. Bloch 1932, Bubeník 1995); -es is not unknown to Sinti varieties, e.g. Piedmontese Sinti witnessed by Annibale Niemen (1995) shows forms as d-el-es 'he did'. The second part of the Shinto Rosengro remoteness marker, -ta, is possibly connected to the morpheme -tar, found in many eastern varieties in combination with verbs, and completely unattested in Western and Northwestern Romani varieties up to now. As Hancock 1995 underlines, in many Vlach varieties the morpheme -tar is added to inflected verbs of motion to express the meaning 'off ', 'away', es. našélas-tar lestar 'she was running away from him' (Hancock 1995:101). Nonetheless, its status in Romani is not clear yet, as suggested by Boretzky & Igla 2004: presumably connected in some way to the ablative -tar, some of its uses do not match with the meaning 'away', as it can be found in addition to copula forms, with the verb avel- (cfr. Borezky & Igla 2204, Teil 1: 174) and, according to the data in the Romani Morpho- Syntax Database, with verbs that do not express movement. Noteworthy is the Kalajdži dialect of Montana (Bulgaria), in which the marker -ta(r) is attested bothin the copula and in verbs with value of repetition or durativity, es. seko bevel vov mangel-as-tar te žal khacende 'every evening he wanted to go somewhere', voj phirl-as-ta pala eke mrušuste 'she was walking behind a man'. The paper will explore the possible paths of grammaticalization that led to the development of the remoteness marker -esta in Shinto Rosengro.

-- Bloch, Jules. 1932. Survivance de skr. āsīt en indien moderne. In: Bulletin de la Société de linguistique de Paris 33(1): 55–65.
-- Boretzky, N. & Igla, B. 2004. Kommentierter Dialektatlas des Romani. Teil 1: Vergleich der Dialekte; Teil 2: Dialektkarten mit einer CD-ROM. Wiesbaden: Harrassowitz Verlag.
-- Bubeník, Vít. 1995. On typological changes and structural borrowing in the history of European Romani. In: Matras, Y. ed. Romani in contact. The History, Structure and Sociology of a Language. Amsterdam/Philadelphia: John Benjamins Publishing Company: 1-24.
-- Caccini, S. a c. di Barontini, M. e Piasere, L. 2001. La lingua degli Shinte rosengre e altri scritti. Roma: CISU.
-- Hancock, Ian. 1995. A handbook of Vlax Romani. Columbus: Slavica Publishers, Inc.
-- Matras, Y. 2002. Romani: A linguistic introduction. Cambridge: Cambridge University Press.
-- Niemen, A. 1995. O ker kun le penijà. La casa con le ruote. Roma: Sinnos Editrice.
-- Romani Morpho-Syntax Database, http://romani.humanities.manchester.ac.uk/rms/


On one unusual function of a sociative adverb in Latvian Romani toup

Natalia Perkova
Stockholm University

In Latvian Romani, the understudied Romani variety spoken in Latvia, Estonia, and Northern Lithuania, there is a set of constructions characterised by a rather remarkable use of the sociative adverb khetané 'together', see (1-4). This type of constructions has a direct parallel in Latvian and various Finnic languages, where dedicated adverbial elements are used as non-basic comitative markers in a number of syntactic patterns and with a restricted set of predicates, cf. (5-6). These markers are glossed as presuppositional comitatives (PCOM) as one of the co-participants forms part of the presupposition.

In my talk, I will present the data illustrating the use of this comitative construction in Latvian Romani against the background of the Circum-Baltic languages Latvian Romani speakers have been exposed to. The Latvian Romani examples have been mainly extracted from the previously unpublished collections of 16 texts collected by Jānis Leimanis in Latvia and Paul Ariste in Estonia in the 1930s; the complementary data come from the Romani Morphosyntax database and the modern collection of Latvian Romani fairytales, as well as from several publications by Paul Ariste.

In addition, this pattern of contact-induced development will be compared to the direct matter borrowing in the Romani dialects in strong contact with German. More precisely, the German prefix mit- used in the constructions in focus is more desemanticised and at the same time formally identical to the comitative preposition mit. The Circum-Baltic dedicated markers, in contrast, are less bleached and are rather functionally aligned to the sociative adverb 'together' due to the similarity of the comitative and the sociative constructions. The existence of an adverb used as a sociative marker in Latvian Romani makes it possible. The analysis shows that the expansion of the adverb khetane in Latvian Romani complements the existing knowledge of contact-induced change in the comitative domain in European Romani dialects, and at the same time points at the salience of this comitative construction in the Circum-Baltic languages.



-- Ariste 1973 = Ariste, Paul. 1973. Einige Märchen der Cuchnú-Zigeuner. Töid Orientalistika Alalt / Oriental Studies 2: 5-40. (Acta et Commentationes Universitatis Tartuensis 309).
-- Ariste ELMM = Paul Ariste. Eesti läti mustlaste murdest (manuscript). Available at: https:// fennougrica.kansalliskirjasto.fi/handle/10024/87067 (the Zingarica collection by the National Library of Finland)
-- Jānis Leimanis' Latvian Romani Folklore Collection: http://garamantas.lv/lv/collection/620/Jana- Leimana-ciganu-folkloras-vakums
-- Latvian Romani fairytales: http://www.pasakas.net/pasakas/citu_tautu_pasakas/ciganu-pasakas/


The inflection of the imperfect in Abruzzian Romani and its historical background toup

Andrea Scala
State University of Milan

The unusual inflection of the imperfect tense in Abruzzian Romani has been repeatedly pointed out. Giulio Soravia speaks of an "innovative and atypical formation" (Morelli-Soravia 1998, 184), Norbert Boretzky and Birgit Igla of the "most uncommon formation" (Boretzky-Igla 2004, I 171) and Yaron Matras and Viktor Elšík of "great complexity" (Elšík-Matras 2006, 192). Taking as an example the verb ker- "to make" the inflection of the imperfect in Abruzzian Romani behaves as follows: 1sg.kerásana, 2sg.kerésana, 3sg.kerésana, 1pl. kerásana, 2pl.keréndza/keréndzana, 3pl.keréndza/keréndzana. As correctly suggested by Matras and Elšik the morpheme -sana, with its post-consonantal allomorph -dzana, can be compared with the sequence -sine, that we find, for example, in Prizren Arli, cfr. kerela-sine 's/he was doing', and originates from the 3sg. of the past tense of the copula, which has been grammaticalized as a remoteness marker.

The process of grammaticalization of the copula in the past 3sg. as a morpheme indicating remoteness is ubiquitous in Romani (Bubeník 1995, 6-10; Matras 2002, 152-155), but in the case of Abruzzian Romani there are two morphs -sana/-dzana and -dza both different from the more common forms such as -as, -a, -e etc. The paper will propose a diachronic reconstruction of the inflection of the imperfect tense in Abruzzian Romani, focusing especially on the 2pl. and 3pl. which show the grammaticalization of sas, another form of the copula, possibly to be traced back to Early Romani as a remoteness proto-morpheme.

Moreover it will explore the hypothesis that Abruzzian Romani preserves some traces of a historical phase in which the morpheme of the imperfect tense was still identifiable in synchrony with the 3sg. of the past of the copula, so that it could be replaced by other variants of the copula. Finally the paper will take into consideration some possible morphological models for the formation of the imperfect offered by languages which have been in contact with Romani. It is not improbable that bilingualism with other languages can have triggered or speeded up the grammaticalization of the copula as a remoteness marker. The case of Kurdish has been already suggested (Matras 2002, 154), but the most striking structural similarity is perhaps that with some Armenian dialects spoken in Northern Iran (e.g. those of Maragha and Xoy, cfr. Adjarian 1909, 81-89; Greppin-Khachaturian 1986, 248-249).

-- Adjarian H., 1909: Classification des dialectes arméniens, Paris, Champion.
-- Boretzky N.-Igla B., 2004: Kommentierter Dialektatlas des Romani, I-II, Wiesbaden.
-- Bubeník V.,1995: On typological changes and structural borrowing in the history of European Romani, in Y. Matras (ed.), Romani in contact. The history, structure and sociology of a language, Amsterdam/Philadelphia, John Benjamins Publishing Company, pp. 1-24.
-- Elšík V.-Matras Y., 2006: Markedness and Language Change, Berlin-New York, Mouton de Gruyter.
-- Greppin J.A.-Khachaturian A. 1986: A Handbook of Armenian Dialectology, Caravan Books, Delmar (New York).
-- Matras Y., 2002: Romani. A Linguistic Introduction, Cambridge, CUP.
-- Morelli B.-Soravia G., 1998: I pativ mengr. La lingua e le tradizioni dei rom abruzzesi, Roma, Centro Studi Zingari.


Three-participant events in Romani toup

Melanie Schippling
University of Cologne

This talk presents the first results of an investigation into the morphosyntactic expression of three- participant events in Romani. It provides an inventory of attested strategies and situates Romani within a typological framework. Events that involve three participants are cross-linguistically often associated with ditransitive constructions that express transfer events such as, for instance, the English verb 'to give'. However, the analysis of three-participant events to be presented here is not limited to these construction types. A broader understanding of three-participant events has been developed by Margetts and Austin (2007). In their cross-linguistic survey, Margetts and Austin (2007: 393) emphasize the existence of "a variety of constructions that are syntactically two-place but express a third participant by some other means". Following their account, three-participant events are understood as "dynamic states of affairs that crucially involve three entities in their conceptualization" (ibid.: 397). Encoding means identified by Margetts and Austin (cf. 2007: 402-403) include three- place predicate strategies, oblique and adjunct strategies, and several others. Their typology of three- participant events constitutes the framework for the analysis presented in this talk.

This talk provides an inventory of the strategies attested in Romani, thereby enabling further research on what happens to a basic inventory when exposed to (intense) language contact. A selection of one focus dialect per Romani dialect group is made in order to investigate encoding strategies in a feasible way: Kalderaš (Vlax), Arli (Balkan), East Slovak Romani (Central), Sinti (Northwestern), and Lithuanian Romani (Northeastern) are considered primarily in this talk. The data set that the analysis draws on consists of items of the RMS questionnaire and phrases in grammatical descriptions of the respective dialects. This means that the findings are mainly based on elicitation data.

My research indicates that three-place predicate and oblique/adjunct strategies are the principal means for encoding three-participant events in Romani, each applicable for several event types. However, differences between the dialects become apparent: Compared to the other dialects, Lithuanian Romani shows a clear preference for the direct-argument strategy, while the other four focus dialects show more variation between the direct-argument and the oblique/adjunct strategy. A causative strategy appears to be productive only in dialects exposed to contact with languages containing productive causative morphology. For the oblique/ adjunct strategy, speakers draw on the Layer II cases ablative, instrumental, and locative, or on inherited and borrowed prepositions. Due to a general retention of synthetic case marking, Lithuanian Romani often shows Layer II cases where other dialects use prepositional phrases. Other strategies are either used less frequently or not found in the data and will only be considered briefly.

-- Margetts, Anna & Peter K. Austin. 2007. "Three-Participant Events in the Languages of the World: Towards a Crosslinguistic Typology."

45(3). 393-451.
-- Matras, Yaron & Viktor Elšík et al. 2001-2018. "Romani Morpho-Syntax Database." URL: http://romani. humanities.manchester.ac.uk/rm/ (21.01.2018).


Progress report on the Dialectological Atlas of Northern dialects toup

Anton Tenser
University of Helsinki

Scholarship on Romani dialectology exhibits some tension in the approaches used to analyze similarities and differences across the dialects, namely the genetic vs. the geographical diffusion model. This tension translates into divergent tactics applied to the tasks of describing and especially classifying Romani varieties.

The older accounts have traditionally relied on the genetic model. The model interprets dialectal variation by proposing that a single variety splits into several varieties, which grow further apart with time, and then divide further in the same fashion. This is the assumption behind the classifications of Romani dialects by Bakker (1999), Boretzky (1999; 2000a; 2000b), and extensively in Boretzky and Igla's (2004) atlas of Romani dialects. A competing, geographical diffusion model has been proposed by Matras (2002). The model asserts that the variation among dialects is subject to the geographical continuums and is best accounted for by assuming linguistic contact between the speakers of adjacent territories. This model has been applied to the NE group in Tenser (2008).

Using the RMS questionnaire, the Helsinki Romani project has collected close to a 100 samples of Romani spoken around the Baltic Sea area, with a view to discern the validity and applicability of the two approaches. The granularity of this data set of spoken Romani varieties is rivaled only by the wealth of Central Romani data collected by the micro-dialectology effort conducted at Charles University, Prague. The work on the Dialectological atlas of Romani spoken around the Baltic Sea area, a volume resultant from the data collection efforts by the Helsinki Romani project, is well under way. The paper will share some general remarks on the methods used to present feature variation in the atlas, as well as some preliminary findings that the data reveals vis-à-vis the two competing dialectology models.


Romani elements in the jargon of Italian funfair: Power and prestige relationstoup

Chiara Tribulato
Ca' Foscari University of Venice

The jargon of the marketplace, commonly known as Dritto and first thoroughly described by Menarini in 1959, is considered one of the most widespread and ancient descendants of the Italian historical jargons documented since the start of the 16th century.

Dritto stands out because it is an in-group secret language that used to hold together vagabonds, artisans, musicians, magicians, peddlers, animal trainers and other members of the peripatetic community, who travelled through northern Italy following seasonal markets and traditional fairs. As soon as this context broke apart over the centuries, the jargon suffered a great shrinkage both in speakers and vocabulary, but it is to this day still in use within the somehow shielded worlds of circus and funfair, due to their direct lineage back to fairs.

The contact between this marginal context and the gypsy world is as frequent as it is unexplored. Both travelling entertainers and sinti communities have been involved in fairs in the same areas since the 18th century, far before circus and funfair were defined as autonomous contexts. Nowadays, the funfair is a complex system where non-gypsy travelling entertainers and Italian sinti share the same living and economic space.

Furthermore, even though Dritto - like other special languages - has been influenced over time by the contact with Romani language (more precisely, by Lombard Sinti), my field research has nonetheless shown that in some areas a completely different process of substitution is under way: over the course of just one generation, the vocabulary of the historical jargon has disappeared almost entirely and has been substituted by sinti loanwords. Therefore, what now seems like a linguistic continuum – within which speakers can choose in every instance one variant rather than another – is arguably becoming a new Romani-based jargon, something completely different from the past.

In conclusion, to understand these changes and how speakers perceive them, provides us important insights into power relations and the growing - though sometimes concealed - prestige of sinti entertainers within the funfair's economic niche.

-- Menarini A.(1959), Il gergo della piazza in R.Leydi (a cura di) La piazza. Spettacoli popolari italiani, Gallo grande, Milano.


From «Paradigm Function Morphology (PFM)» to NooJ morphology toup

Masako Watabe
Sorbonne University

Today the importance of written communication by information technology (e.g. e-mail, chat) is not negligible. According to «Universal declaration of linguistic rights» (Barcelona, 1996), «In the field of information technology, all language communities are entitled to have at their disposal equipment adapted to their linguistic system and tools and products in their language». However, Rromani speakers do not fully benefit from the development of new technology (e.g. spell checkers, electronic dictionaries) in their own language until now. To develop an IT tool, a simple dictionary of words is not enough as linguistic resources, but a morphosyntactic study would be required.

First, I will introduce the modelisation of inflectional paradigms through «Paradigm Function Morphology (PFM)» (Stump 2001, Bonami & Stump 2016) in order to establish a coherent inflectional morphology from a computational point of view. Secondly, I will apply this morphology to the NooJ module for the Rromani language, a computational system including dictionary and grammar (morphology and syntax). NooJ grammar is applied to NooJ dictionary to generate automatically all inflected forms and to annotate a text.

PFM is an inferential and realisational theory, as opposed to lexical theories. According to PFM, each morphosyntactic feature is associated with a rule that conditions the final form, and each marker can be realised only in a form already associated with these features. For example, ćhavo Rromani boy, son and raklo non-Rromani boy are part of an identical inflectional class (CF) as well as other masculine human nouns with a final vowel «-o», on the other hand, kàko uncle and doktòro doctor are not part of this CF, yet they are also masculine human nouns with a final vowel «-o». In fact, the CF applicable to ćhavo and raklo requires another condition; it is oxytonic. The opposition «oxytonic vs. non-oxytonic» conditions the entire nominal morphology of Rromani. Then, non-oxytonic nouns will be divided with other morphosyntactic features.

Indeed, this reflection is very close to preparatory work for computer programming. PFM theory allows to clarify and to consolidate the NooJ morphology.

As lexical and morphological data, a polylectal dictionary (i.e. including variants of the main dialects and vernaculars of the Rromani) is imported into the NooJ dictionary for the Rromani. The objective of creating a polylectal module is to contribute a pragmatic tool of the written communication to the entire Rromani speakers while keeping their diversity of dialects and living places.

This project is developed encompassing the various dialects of Rromani and is aimed among others at coining softwares enabling the transfer from one dialect into others. Then, users of the Rromani module could immerse themselves in the dialectal diversity of this language even without recognizing it and without losing it. This fact could justify the similarity of the Rromani dialects and the inseparable unity of this language while preserving diversity as a cultural heritage.

Key words
Rromani language, morphosyntax, automatic language processing, NooJ, PFM.

-- Bonami O. & Stump G. T, «Paradigm Function Morphology», In Hippisley A. & Stump G. T. (Edit), The Cambridge Handbook of Morphology, Cambridge University Press, Cambridge, 2016.
-- Courthiade M. et al., Morri angluni rromane ćhibaqi evroputni lavustik (My first European dictionary of the Rromani Language), Cigány Ház, Budapest, 2009.
-- Silberztein M., La formalisation des langues : l'approche de NooJ, ISTE, London, 2015.
-- Stump G. T., Inflectional Morphology: A theory of paradigm structure, Cambridge University Press, Cambridge, 2001.
-- NooJ web site: http://www.nooj-association.org/
-- Universal declaration of linguistic rights, Barcelona, 1996.


Working with linguistic data – experiences from the Norwegian Romani project toup

Jakob Anton Paul Wiedner
University of Graz

After the Norwegian Romani project at the University of Oslo has been finished in 2017, I would like to present the project outcomes from a computational and technical point of view. The aim of the project was to shed light on the way how Norwegian Romani, the varieties of so-called Scandoromani spoken in Norway, is discursively constructed in both scientific and non-scientific literature, in government white papers and in interviews with speakers of Norwegian Romani. Although it is thus not a 'classical' linguistic project, I was able to gather structural linguistic data on this Para- Romani variety both in published sources and through interviews with the speakers. As a result, I was confronted with data that were partly publicly available, partly on-line available, partly only in books, while the data collected during the project were available only to me. In order to create a knowledge base that is available to the Norwegian Romanifolk/Tater community but also to the scientific community, I decided to publish the lexical data gathered in an on-line dictionary (app.uio. no/hf/nro) and to write an article on morphological particularities accompanying the use of Romani vocabulary used in Norwegian discourse (Wiedner 2017).

In my presentation, I want to present my approach to setting up a database for the documentation of Norwegian Romani lexicon and to publishing it as a dictionary that can be accessed as via a website. First, I will describe the database design and I will explain my choice coming from both linguistic and computational considerations. Thereupon, I will present the digitalisation process with special attention to possible dangers of entering incorrect data and ways to evade such problems. Finally, I will speak about the difficulties when publishing linguistic data on-line such as questions of appropriate website design, comprehensibility of the website, i.e. the question if the target group will be comfortable with the way the website works, but also security issues.

My talk is intended to help researchers who are currently planning projects aiming at digitalising and publishing the research findings in the internet, but also to invite other researchers within Romani linguistics to exchange their knowledge and experiences with regard to computer-assisted language data gathering and processing.

-- Wiedner, Jakob A. P. (2017): Norwegian Romani – an integrational framework. In: Kozhanov, Kirill, Oslon, Mikhail and Halwachs, Dieter (eds.): Das amen godi pala Lev Čerenkov. Romani historija, čhib taj kultura. Graz: Grazer Romani Publikationen, pp. 455-483.




Imprimer Contacter le webmestre Plan du site Crédits Accueil