Author: Urech, Evelyne; Van Den Heuvel, Wilco
Date published: December 1, 2011
Journal code: RMNS
A note on spelling
Group names will be given in the plural form throughout the article. For all group names based on Romanian words, the Romanian spelling is used (e.g. Cäldärari, Bätest, Läutari). Group names based on words from languages other than Romanian (but which are used as well when speaking Romanian) are dealt with analogously (e.g. Lovari, Calaposi, Cantale). For Romani words, and any other group names, the usual academic transliteration is used (e.g. kast, roma vlaxika, Cerhari, Burgudzi).
Table 1 lists the equivalent of several sounds in different writing systems. Strict phonetic notation (IPA) and the official orthography for Romani dialects spoken in Romania are not used in this article but rendered here for the ease of use for readers more familiar with one of these transliteration systems.
The research presented in this article forms only a part of a broader, ongoing survey of Romani dialects and the sociolinguistic situation of Roma in Romania. The project stands under the aegis of SIL International and is carried out by a team of researchers affiliated with this organisation.1 Part of the research tools and valuable feedback was provided by Yaron Matras of Manchester University, who was also involved in planning the survey. Field work was divided into two periods: a period of pilot testing done in the Mureç County in May and June 2007; and a phase of extensive data-gathering all over Transylvania between July 2007 and June 2009. Survey team members visited 111 Roma communities, collecting 135 short word lists, 21 longer word lists, and conducting 131 sociolinguistic interviews, through the medium of Romanian or, in a few cases, Hungarian.
This article addresses one specific aspect of the survey, i.e., Roma group names. A thorough report of the full survey will be published in the SIL International electronic survey report series (http://www.sil.org/SILESR/).
1.2.1. Research tools
The survey draws on the following research tools.
18.104.22.168 Word lists
Two word lists were used in order to identify local language varieties of Romani. One of them comprises of 32 words and short phrases, containing some of the key diagnostic features of Romani dialect variation. The initial version of this list was provided by Yaron Matras from Manchester University and has been slightly elaborated by the survey team in the course of the research. Respondents were asked to translate the 32 items from either Romanian or Hungarian into their variety of Romani. All the answers were electronically recorded and transcribed on paper on the spot using the IPA notation. Later, recordings were listened to again to check for things we might have missed. The use of this short word list enabled us to efficiently classify local varieties of Romani, while also minimising any hesitancy of respondents to participate in the survey as it was not too time-consuming. In a few cases (twenty locations)2 an extended word list (containing 1060 words, phrases, verb paradigms, and a short narrative) which had been developed by the Romani Project hosted at Manchester University was also collected. This long word list was only elicited in Romanian, thus Roma respondents translated orally from Romanian into their variety of Romani, while we electronically recorded their answers. Based on these recordings, all entries were transcribed employing the conventions used in Romani linguistics, either by the staff of Manchester University, or by survey team members Wilco van den Heuvel and Sarolta Gardner. The long word list provides a deeper insight into several local varieties of Romani.
22.214.171.124. Sociolinguistic interviews
In nearly all instances the short word list was followed by a guided interview about the personal background of the respondents, the local Roma community, language use patterns, dialect intelligibility, education, etc. The researchers had a set of questions and topics in mind which they covered during an interview, however, the respondents were allowed the freedom to add other topics they found relevant. While the researcher had one main conversational partner who provided the information for the word list, many interviews took on the character of a group discussion, which helped to get a multi-faceted insight into the subjects brought up. All interviews were done either in Romanian or Hungarian, and were recorded.
Observation of natural language use revealed much about the vitality of the language, and it helped in evaluating to which degree the reported language use practices correspond with the actual linguistic behaviour in the community.
Sampling was not done in a strictly statistical way but rather by intensive networking. Contacts with the local Roma communities were established through contact persons (both non-Roma and Roma) from various NGOs, village or town administration, churches, the Roma political party, school teachers, medical doctors, etc. In addition to these more 'official' ways of being introduced, several respondents provided the researchers with addresses of family members or acquaintances whom they considered as being able to contribute interesting data to the present study. Thus, as time went by, the network of contacts grew <organically>. The data collected in this way is valuable and can be considered representative for various reasons, the mains ones being (a) the good geographical coverage and the large number of places visited; (b) contacts were made through representatives from very different social networks; and (c) new data confirming what has already been found.
2. Basic linguistic findings
When we compared the short word lists, it became evident that there are four linguistic factors which regularly correlate. These factors are:
a. s vs. h in different positions (e.g. instrumental case ending (-5a, -ha), copula (som, horn))
b. s vs. c in certain lexical items (e.g. savolcavo 'boy')
c. 'nothing' with vs. without -n- (e.g. khanc(-i) vs. khajc(-i))
d. 'tomorrow' related to teha-ra)ltehe or to tejsaitejse
On the basis of the correlations that we found, local language varieties in our sample can clearly be divided into two categories. As presented in Table i, half the varieties are characterised by the correlation of s, s, a khanc(-i) related form in the word for 'nothing' (indicated in the fourth column by a +), and a teha(ra)itehe related form for 'tomorrow'. As these varieties clearly dominate in the southern part of Transylvania, in this article we will refer to them as South Transylvanian (ST) dialects. Another 40 per cent of the local language varieties in our sample show the correlation of h, c, a khajc(-i) related form in the word for 'nothing' (indicated in the fourth column by a -), and a tejsaitejse related form for 'tomorrow'. We will refer to them here as North Transylvanian (NT) dialects, as the local language varieties making up this group prevail in the north and north-east of the area under study. Only 10 per cent of all the local Romani varieties examined did not fit neatly into one of the two categories. Further study is needed in order to evaluate to which degree this is due to insufficient or deficient data, or to inter-dialectal contact.
It should be stressed that the terms NT and ST are used here simply as descriptive labels to refer to the two categories which are established empirically. In this article we will show, however, that these two categories can be correlated to certain group names, and as such reflect certain historical developments.
In the southern Transylvanian counties (Alba, Arad, Braçov, Huneadoara, Sibiu, and Timiçoara) and in the east (Harghita) only ST dialects are found. In the two northern Transylvanian counties, Maramureç and Bistrita-Näsäud, NT dialects clearly dominate. In all other counties both groups are represented with no detectable geographical distribution between them. This means that speakers of different Romani varieties live closely together, sometimes even in the same village. In a situation like this it must be expected that the different dialects influence each other. But even though NT and ST dialects have a lot in common, they can still clearly be distinguished - based on the four factors presented above.
We plan to publish the in-depth linguistic findings, data about intra-group variation, and a discussion about the position of the Romani varieties present in Transylvania with regard to the Romani landscape in general. For the time being, we will limit ourselves to stating that - in very general and simplified terms - the 'North-Transylvanian features' are either exclusive to the area (khajc(-i)), not described elsewhere, or similar to certain varieties found to the north and west of Transylvania (tejsaltajsa, h in instrumental or copula) (Boretzky and IgIa 2004a: 198; Matras 2002: 229O, while the 'South-Transylvanian features' (tehara, khanc/khanci) are more characteristic of Vlax as a whole (Boretzky and IgIa 2004a: 108, 198, Matras 2002: 224) or Northern Vlax specifically (s instead of c) (Boretzky and IgIa 2004a: 250).3
For the purpose of this article it suffices for the reader to be aware of the existence of these two broad dialect groups. In the following section we will set out to show that there exists a strong correlation between these two groups and certain group names.
3. Group names
3.1. General remarks
The Roma are characterised by a great internal heterogeneity. Subgroups are differentiated by various factors such as life style, religion, language, internal structures, period and routes of migration, and level of integration in the broader national community. The names used to refer to the subgroups often reflect these factors while the actual terms Roma use are generally borrowed from Romanian, Turkish, or Hungarian (see Bakker et al. 2000: 6of, Matras 2002: 5f, Tcherenkov and Laederich 2004: 275-309):
- Names deduced from (traditional) occupations, e.g. 'horse dealers': Lovari (Hung. Io 'horse'), 'drill makers': Burgudzi/ Bugurdzi (Turk, burgucu gimlet maker'), 'comb makers': Pieptanari (Rom. pieptene comb'), 'musicians': Lautari (Rom. làuta 'lute, stringed instrument').
- Names related to present or former place of residence, e.g. Bergicka Roma (Germ. Berg 'mountain')4, Burgenland Roma (Burgenland is the easternmost Austrian province), Macvaja (Malva is a region in northwestern Serbia).
- Names relating to a common ancestor, e.g. Gabon, descendants of a clan leader ofthat name.
- Names conveying religious affiliation, e.g. 'Muslim Roma: Xoroxane (Romani xoraxajlkoraxaj 'foreigner', or ultimately from Qur'an).
- Names referring to dominant surrounding people groups, e.g. 'Hungarian Roma, 'Polish Roma.
- Names relating to way of living, e.g. 'settled people: Arlija (Turk, yerli 'local'), Rromi de vatrâ (Rom. vaträ 'hearth'), 'Tent Roma: Cortorari (Rom. cort 'tent').
- Names marking social standing, e.g. 'Silk Gypsies': Tigani de matase.
Group names are not static concepts, but a means used by individuals, or groups of people, to position themselves with respect to other individuals or groups. Thus a single group of Roma can use several endonyms, depending on how it wants to present itself, and at the same time this group can also be given different exonyms, depending on either the person using it, or on the context in which it is used. Moreover group names can be a reason of debate or negotiating:
Roma placing themselves in one of the main groups do not necessarily accept all other Roma placing themselves in the same main group. In other words, some Roma who declare membership of a group are not regarded as belonging to the group by others in the same group. (Szuhay 2005: 237)
Even though group names are generally flexible and can change according to the situation, or over time, there is often a tendency to conservatism in name-giving. Because of this, certain names remain in use even after the conditions that led to the application of the name have disappeared (e.g. Ursari, from Rom. urs 'bear', are no longer 'bear trainers', most Lovari don t work with horses anymore).
Group names with the same semantic meaning, like for example Ursari, Meckara (Serb, mecka 'bear'), Medvedara (Slovak medved' 'bear'), Rickara (Romani rie 'bear'), and Ajdzijes (Turk, ay 'bear'), do not necessarily imply that all the Roma bearing those names are culturally and linguistically very similar, some of them might not even speak Romani anymore (Hübschmannová 2003, Bakker et al. 2000: 60).
One case study will suffice to illustrate here how complex subgroup labelling can be. In Deaj (Mure§ County), the Romani speaking community refers to itself as (1) Cärämidari (Rom. caramida 'brick') 'brick makers', indicating their traditional profession, but also as (2) Tigani de casa 'House Gypsies', to express their position as settled Roma in opposition to the Cortorari 'Tent Roma, and as (3) Tigani Românizati 'Romanianised Gypsies', to express a closer relationship to the Romanian population and to distinguish themselves from more "traditional", less assimilated Roma, and finally also as (4) Tigani de matase 'Silk Gypsies', thus declaring their way of living as more "refined" in comparison with other Roma groups. The Cortorari 'Tent Roma, on the other hand, refer to the Roma from Deaj as (1) Romunguri, a term that could be translated as 'Hungarian Roma and which is used to express their (past) assimilation to the Hungarian population (cf. the section on Cortorari and Romungri below) which again is presented as standing in contrast to the Cortorari s own more pure and "original" position, or in Romani as (2) LoIo po pdr '(Those who are) red on the belly',5 a Romani expression which seems to be used as a mocking name for groups that are considered Romunguri.
3.2. Data collection
The information on Roma group names presented here is based exclusively on what Roma people say about themselves and about others. The relevant topics discussed with respondents during the sociolinguistic interview were: a. selfdesignation, and b. designation given by other Roma. We thus received four types of group names: endonyms ('We are Tigani de mâtasa!), exonyms ('They are Tigani de mâtasa.), reported exonyms ('They call us Tigani de mâtasa.), and reported endonyms ('They say that they are Tigani de mätasa). Although interviews were conducted in Romanian or Hungarian,6 in a later phase of the research the group names were asked for using a Romani dialect the researcher knew. Additionally respondents were asked about (traditional) trades, ways of living, internal community structures, customs (marriage, clothing), religious appurtenance, dialect intelligibility, language use patterns, etc.7
For the present analysis only ethnonyms which were mentioned to the researchers are taken into account. This type of data is subjective in nature, but provides significant insights into the name-giving processes at hand. It will be correlated with the information on the local language varieties (as obtained from the short word lists), which could be considered as data which is less prone to subject bias.
3.3. Data discussion
3.3.1. No specific name
Concerning names for Roma subgroups, the first observation to be made is that about 45 per cent of all respondents did not associate with any specific subgroup, but instead only gave the generic Roma (in Romani) or Tigani (in Romanian) as an answer to the question about what type of Roma they were. This pattern of self- reference is mentioned elsewhere too (Bakker et al. 2000: 61, Matras 2002: 5, Tcherenkov and Laederich 2004: 277). In our case, respondents were familiar with typical group names, but they often mentioned them only in order to state that their group does not belong to any of the clusters mentioned.
3.3.2. Status of the term 'Gypsy'
The second observation to be made is that the term Gypsy (Tigan in Romanian, Cigány in Hungarian) is widely accepted and used in a neutral manner when conversing in the majority language. The only people we met who explicitly preferred the use of Roma or Rromi over 'Gypsies', considering the latter derogative, were trained teachers of Romani language and culture, or members of the Roma Political Party. Otherwise people generally preferred to use the term Gypsy, saying: "We are proud of being Tiganir, "We are not ashamed of being Tigani!", "Only the ones 'up there' in politics insist that we should be called Rromir, or they were indifferent with respect to the use of either of the terms.
3.3.3. Relations between ethnonyms and dialects
The third observation to be made is that a correlation between dialects and group names can be established. In the following paragraphs we will elaborate on this point while restricting ourselves to some of the most commonly attested group names.
126.96.36.199. General division into two groups
Figure 2 shows a basic division of Roma in Transylvania into speakers of ST dialects (under the broken line) and speakers of NT dialects (above the broken line). The two circles are representative of certain group names, while the size of the circles is a rough and intuitive indication of the number of Roma for whom this term is, or could be, used.8
The group names given in this figure are mutually exclusive and each one is associated with only one of the two dialect groups. The Romanian term Cortorari/Corturari 'Tent Roma (or in Romani Cerhari, from cerha 'tent')9 is used exclusively for speakers of ST dialects, while the name Romungri/ Romunguri/ Rumungri/Rumunguri/Rromi Unguri/Romungurisa/Romunglâi is used exclusively for speakers of NT dialects.
Compared to other group names used for speakers of ST dialects, Cortorari is the most frequently applied name. Whereas it is almost always used both as a self- designation and as an outsider-designation, we have met some cases where it was only given either as an exonym ("They are Cortorari"), or as a reported exonym ("Others call us Cortorari?). In two instances, both in Cugir, Alba County, respondents stated that others call them Cortorari, but they themselves explicitly denied this. The term Cortorari 'Tent Roma historically points to (semi-)nomadic groups. Nowadays, the name seems to be associated with certain features which are considered as traditional, like traditional dresses (hats for men, long skirts and long hair for women), certain marriage customs, the presence of group leaders (buliba§ä), or the traditional court (romani kris).
Contrary to Cortorari, the name Romungri is used exclusively as an exonym. The singular form rom ungur or rumungur could be translated as 'Hungarian Rom' and is probably an expression of the fact that historically Romungri lived among Hungarians, and would typically use Hungarian as their first second language, i.e., the first language learned after or next to Romani. Although many Romungri would refer to themselves as 'Hungarian Roma, (in Romanian: Tigani Ungure§ti, Tigani Unguri), it should be noted that the present use of the term 'Hungarian Roma in Transylvania does not fully coincide with the notion of Romungri, a phenomenon which we will come back to below. Moreover, in certain regions the name Romungri is used for Roma who no longer know Romani.
The clear correlation between these two ethnonyms and the respective dialect groups is a strong indication that the two sets of correlating features, which we termed as ST and NT dialects, are not just a theoretical construct but indeed reflect a historical reality. Speakers of NT dialects seemed to have been settled in Transylvania for a longer period of time and typically had Hungarian as their first second language. In the course of time, they assimilated to a certain degree into the majority population, losing many of their traditional customs. Some even gave up speaking Romani. This line of reasoning conforms to the fact, often pointed out, that the term Romungri is used for groups who settled down long ago (see Hübschmannová 2003, Bakker et al. 2000: 61). The speakers of ST dialects, on the other hand, led a nomadic or semi-nomadic life until quite recently, and generally did not assimilate as much into the majority population as the Romungri.
188.8.131.52. Groups speaking South Transylvanian varieties of Romani
Now, focusing only on group names used by, or for, ST dialect speakers, one can extend the lower part of Figure 2 as in Figure 3.
As can be seen in Figure 3, some communities of ST speakers are referred to as Ciurari 'sieve makers' (Rom. ciur 'sieve'), indicating their historical profession. This term is used both as an endonym and exonym, as is the case for Cortorari 'Tent Roma, while for some of the Ciurari communities the name is acceptable only as an exonym. Some of the Ciurari would also be considered Cortorari (by themselves or by others).
Another group to be mentioned are the Cäldärari (also Kalderash/Kelderash) or 'kettle makers'. The name indicates a historical profession and is used both as an endonym and exonym. Although, in our interviews, none of the Cäldärari spontaneously mentioned Cortorari 'Tent Roma as an alternative label for their group, some respondents did accept this name as an appropriate designation after having been asked about it.
Finally, there is the group which is known under the name Gabon. Attested exonyms for this group are Cortorari 'Tent Roma, CiotârnariiCetârnari (Hung. csatorna gutter, drainpipe), referring to their profession of guttering, same as badogosi (Hung, badog'tiri). Because of the hats worn by Gabori men, the group can also be referred to as Calaposi (Hung, kalap 'hat'). This term is accepted by Gabori as a self-appellation as well. The highest concentration of Gabori is found in predominantly Hungarian areas. This explains why they refer to themselves, and are referred to by others, as Tigani Unguresti/ Tigani Unguri 'Hungarian Roma',10 a label otherwise not used for ST speakers (see next section). Some writers attest to an overlap between Gabori and Cäldärari 'kettle makers' or Ciurari 'sieve makers' (Berta 2007: 32, Olivera 2007: 11). On the basis of our data, this cannot be evaluated because there are very few instances of respondents mentioning these names in combination. The language variety which the Gabori speak can clearly be classified as a ST dialect. But compared to other ST varieties, the Gabori variety forms a separate subgroup which is distinguished not only by linguists (e.g. Gardner & Gardner 2008), but also by Gabori themselves and by the speakers of other Romani varieties.
184.108.40.206. 'Romanian Roma' and 'Hungarian Roma'
Many Roma in Transylvania refer to themselves as either Tigani Românesti/ Tigani Romàni 'Romanian Roma or as Tigani Unguresti/ Tigani Unguri 'Hungarian Roma. One can now add these terms to the figure, see Figure 4.
As stated above, the word Romungri is used as an exonym for certain groups of Roma who speak a NT dialect, and who typically used to live among a majority population of Hungarians, having Hungarian as their first second language. Nowadays, however, the name Romungri does not necessarily imply proficiency in Hungarian, as the name can also be used for groups who have adopted Romanian as a first second language (as in Band or Deaj, both in the Mureç County). While the name Romungri does not bear any relation to the present use of Hungarian, both the Romani designation roma ungrika 'Hungarian Roma and the Romanian terms Tigani Unguresti! Tigani Unguri 'Hungarian Roma'11 do. They are opposed to the Romani terms Roma vlaxika/ Roma laxika/Roma lexika, 'Walachian/Romanian Roma or the Romanian terms Tigani Româneçti/ Tigani Romàni 'Romanian Roma. For ease of description, we will use the English terms 'Hungarian Roma and 'Romanian Roma here. 'Hungarian Roma have Hungarian as their first second language, and they generally feel more connected to the Hungarian population than to ethnic Romanians. 'Romanian Roma are bilingual in Romani and Romanian, and associate more with ethnic Romanians than with Hungarians.
As can be seen in Figure 4, the terms 'Romanian Roma and 'Hungarian Roma are used for speakers of both NT and ST dialects. However, there is a tendency that most 'Hungarian Roma speak a NT variety (the only exception being the Gabori), and many 'Romanian Roma speak a ST variety.
The term Tigani Romanizati 'Romanianised Roma, a term specifically referring to a perceived cultural assimilation into the majority population, was also often encountered, but Tigani Maghiarizati 'Hungarianised Roma was attested to only once (i.e., in Läschia, Maramure§ County).
Roma refer to Non-Roma as Ga(d)ze. Even though this term can be used for all Non-Roma, it is often associated first only with the majority population among which the specific Roma group is living. Thus Ga(d)ze refers to Hungarians in Hungarian dominated regions, and to Romanians in Romanian dominated areas. In certain regions formerly dominated by the Saxons (Germans) Ga(d)ze still refers to Saxons primarily (e.g. Petriç, Bistrita-Näsäud County). Following this, the Romani expression Roma Ga(d)zikane can correspond either to 'Romanian Roma or 'Hungarian Roma, depending on the present or historical situation of the Roma for whom this designation is used.
3.3.4. Some other group names
In this last section we would like to list several common Transylvanian group names which are not associated with either the NT or ST dialect groups.
One of those common group names is Cärämidari (or Cärämizari) 'brick makers', a name which indicates a historical profession.
Another name found among both NT and ST groups is the name Tigani de matase or 'Silk Gypsies'. This term is mainly used as a self-designation. 'Silk Gypsies' are seen as having a higher status, or social position, being more "refined" or more like the majority population in comparison to other Roma groups. Ries describes how the term is used to indicate wealthy Roma who have managed to adapt or even over-adapt to the Gadze world (Ries 2007: 97-106, 212-14).
Finally, we would like to mention the term Castale/ Castaldi 'wood workers' (Romani hast 'wood'), to which some respondents gave the Romanian equivalent bätest/ beiasil beast 'miners, gold washers' (Rom. bates 'person working in a (gold) mine), which is parallel to rudari (Slavic ruda 'metal'), while others mentioned lingurari 'spoon makers' (Rom. lingurä 'spoon') also as a counterpart. Historically all these terms probably refer to settled Roma groups who used to work as miners, therefore the terms Rudari and Batest. At some point many of them left this occupation in order to make wooden utensils (baskets, brooms, spoons, etc.) instead, thus the terms Castale and Lingurari (Saräu 2006: 5). This explains why these four different labels (Bätest, Castale, Lingurari, Rudari) refer, more or less, to the same cluster of people. All of these terms are most commonly used as exonyms to denote Roma who do not speak Romani. However, we found a few exceptions: One respondent from Lugaçu de Jos, in Bihor County, who used Castale to refer to his own Romani-speaking community; and some other Romani-speaking groups who were called Castale! 'Castaldi by outsiders - this was the case in Valcäu de Sus, Sälaj County, and in Ineu, Bihor County. However, in Baia Sprie, Maramure§ County, there lives a community of Roma who speak Romani and use Bätest as an endonym.
This article discussed several group names used by Roma in Transylvania, focusing on the question of how these names are connected to the two Romani dialect groups represented in Transylvania. To conclude, we summarise the significant findings.
Firstly, we found a strong correlation between group names and ST or NT dialects. The fact that the name Cortorari is only applied to speakers of the ST varieties shows a strong indication for the assumption that there is a historical connection between present-day communities speaking a ST variety, and groups of Roma who used to travel around, some of them until recently. At the same time, the fact that the term Romungri is used exclusively for speakers of NT varieties implies a historical continuity between Roma groups who speak NT varieties and groups that are, or used to be, living in an area where Hungarians are the majority.
Secondly, we noted that certain group names are used exclusively for subgroups that speak ST varieties. These names refer either to a historical profession (Ciurari. Câldârari), or to a certain lineage (Gabon). The term Cärämidari refers to a profession as well, but unlike the names mentioned so far, it does not predict whether the community referred to as Cärämidari speaks a NT or a ST variety.
Finally, we found a number of group names which correlate weakly, or not at all, with the two dialect groups. The labels 'Hungarian Roma and 'Romanian Roma are typically used to express the group's first second language, while the term 'romanianised' indicates a perceived assimilation of the group to the Romanian population. The same is true for the term Tigani de matase which is used to refer to groups who have (over-)adapted to the majority population. The names Cantale, Bäie§i and Lingurari are the commonly preferred names used for non-Romani speaking groups.
1. Anna Adámková, B. S. W., licensed social worker (Data gathering); David Gardner, PhysD., physicist (Computer support); Sarolta Gardner (Data gathering); Evelyne Urech, M. ?., linguist & ethnologist (Data gathering & Analysis); Talitha van den Heuvel, M. D., medical doctor (Data gathering); Wilco van den Heuvel, PhD., linguist (Data gathering & Analysis).
2. Albesti I+II (Mures County), Bahnea (Mures County), Branco venesti (Mures County), Cornesti (Mures County), Cuiesd (Mures County), Deaj (Mures County), Diosig (Bihor County), Glodeni (Mures County), Ineu de Cris (Bihor County), Lugasu de Jos (Bihor County), Mäguri (Timis County), Mal (Sälaj County), Näsäud (Bistrita-Näsäud County), Säcuieni (Bihor County), Senereus (Mures County), Spinus (Bihor County), Tämasda (Bihor County), Tinea (Bihor County), Uileacu de Cris (Bihor County), VeIt (Sibiu County).
3. The form tehe is not described as such, nor can it be found in the Romlex online dictionary (http: //rom ani. uni-graz. at/romlex/) .
4. Name used for Roma living in mountainous Poland near the Slovakian border.
5. This expression seems to be used by Cortorari to refer to Romungri. It was applied to the Roma in Senereus and Bahnea (both Mures County) as well. Where the designation stems from is unknown.
6. Only the six interviews done in the Harghita County were conducted in Hungarian. In Harghita Hungarian is clearly the predominant language. However, Hungarian group names are not taken into account in this article.
7. While most of the questions used were inspired by the Manchester Ethnographic Questionnaire (a set of questions elicited in conjunction with the long word list), we did not use their entire list and went into more detail at certain points.
8. We do not have data on the total number of ST speakers nor on the number of NT speakers.
9. See Olivera on the equation of the Romanian term Cortorari and the Romani term Cerhari (Olivera 2007: 157). In our data the term Cerhari is used very rarely.
10. Interestingly, because the Gabori see themselves as 'Hungarian Roma', some have reinterpreted the term Romungri to mean 'Romanian Roma' román-gril
11. Most probably the Hungarian term Magyar Cigânyok is parallel to this.
Bakker, Peter, Hubs dimano vá, Milena, Kalinin, Valdemar, Kenrick, Donald, Ky uchú - ko ?, Hristo, Matras, Yaron, and S or avia, Giulio. 2000. What is the Romani language? Hatfield: University of Hertfordshire Press.
Berta, Peter. 2007. Ethnicisation of value - the value of ethnicity: The prestige-item economy as a performance of ethnic identity among the Gabors of Transylvania (Romania). Romani Studies. Fifth Series. 17 (5): 31-65.
Boretzky, Norbert, and IgIa, Birgit. 2004a. Kommentierter Dialekt atlas des Romani. Teil 1. Vergleich der Dialekte. Wiesbaden: Harrassowitz.
Cortiade, Marcel. 1989. La standardisation du Romanes. In: Williams, Patrick, ed. Tsiganes: Identité, évolution. Paris: Syr os/ Études Tsiganes. 375 -84.
Gardner, David J., and Gardner, Sárolta A. 2008. A provisional phonology of Gabor Romani. Romani Studies. Fifth Series. 18: 155-99.
Hübschmannová, Milena. 2003. Untergruppen der Roma. ROMBASE, http: //ling. kfunigraz.ac.at/-rombase/cgi-bin/art.cgi?src=data/ethn/topics/names.de.xml, accessed 12 Sept. 2011.
Matras, Yaron. 2002. Romani: A linguistic introduction. Cambridge: Cambridge University Press.
Olivera, Martin. 2007. Romanes ou G intégration traditionelle des Gab ori de Transylvanie. Doctorate Thesis. Paris: Université Paris X. (http://www.romeurope.org/ proto/IMG/th ese-martin-olivera-romanes.pdf, accessed 12 Sept. 2011).
Ries, Johannes. 2007. Welten Wanderer - Über die kulturelle Souveränität siebenbürgischer Zigeuner und den Einfluss des Pfingstchristentums. Würzburg: Ergon.
Saräu, Gheorghe. 2006. Dictionar Rrom-Român. Bucharest: Sigma.
Szuhay, Péter. 2005. The self- definitions of Roma ethnic groups and their perceptions of other Roma groups. In: Kemény, István, ed. Roma of Hungary. New York: Columbia University Press. 237-46. (http://www.mtaki.hu/docs/kemeny_istvan_ ed_roma_of_hungary/peter_szuhay_the_self_definitions_of_roma_ethnic_ groups.pdf, accessed 12th September 2011)
Tcherenkov, Lev, and Laedrerich, Stéphane. 2004. The Rroma. Vol. 1: History, language, and groups. Basel: Schwabe.
Evelyne Urech is Sociolinguistic Researcher with SIL International, 7500 W Camp Wisdom Road, Dallas, t? 75236-5629, USA. E-mail: firstname.lastname@example.org. Wilco van den Heuvel is Linguistic Researcher at the Free University of Amsterdam, De Boelelaan 1105, 1081 hv Amsterdam, the Netherlands. E-mail: email@example.com