REFERENCE No.

To cite this output:

RES-000-23-1133
Matras, Yaron (2008). Place, mobility, and dialect differentiation in Romani : Full Research Report
ESRC End of Award Report, RES-000-23-1133. Swindon: ESRC

Research Report

x Background

Romani is the largest minority language spoken in the European Union. Unlike most
other languages of Europe, it lacks a coherent territorial spread, and its speaker
communities are geographically dispersed in socially isolated communities throughout
the continent. Romani also lacks written attestation until very recently. The key to
understanding diachronic development in Romani therefore lies in a comparative
investigation of its present-day varieties. Given the dispersion, the isolation, the absence
of any overarching linguistic norm (or binding social or political framework), and the fact
that speaker communities are always bilingual, variation within Romani is considerable.
The language therefore lends itself as a test-sample for a range of theoretical problems in
linguistics, from the impact of language contact on typology, through to the pace of
structural change and the pathways of internal grammaticalisation developments, on to
the relevance of place and space to the spread of innovations and changes, and finally to
the position that linguistic pluralism can or should occupy when drafting language policy
for Romani today.
The project connects directly to previous work by the PI, much of it in
collaboration with Viktor Elšík (now Charles University, Prague), on the documentation
of dialectal variation within Romani and the reconstruction of historical processes of
change in the language. In particular, it draws directly on a sketch for a database interface
that could present comparative data on the dialects of Romani, taking into consideration
historical developments (form-to-form and form-to-function) as well as the state of the
art in descriptive linguistic typology (function-to-form approach), on a corpus of
preliminary data gathered in earlier investigations, and on a method for data collection
(the RMS questionnaire). As for the framework for analysis, the project draws on
previous work by the PI in functional typology and in particular in adopting a functional
and cognitive-explanatory approach to language contact and convergence phenomena in
language, and on the PI’s work on dialectology and on language codification.
The background for the project is thus a combination of methodological,
theoretical, and practical questions and challenges. The project seeks practical solutions
for the documentation of a language like Romani, which lacks a recognised Standard and
consists entirely of ‘dialects’ of equal status. The project’s approach to this challenge, in
organisational terms, was to put together a committed team consisting of a technical
expert, an archive manager, academics and a series of collaborators in different countries,
who together would develop a strategy to collect and process data and a model for the
storage and presentation of these data. From the onset of the project it was the intention
to share as much data as possible with a wide community of users. The principal concept
governing the project’s strategy was that the use of new technologies can inspire a
revised conceptualisation of data, not just for the practical purposes of archiving and
presentation, but also of the theoretical outlook, in re-visiting traditional notions of
dialect boundaries and dialect branches and thus of constructions that are normally taken
for granted as quasi-realistic discrete entities. Moreover, the project’s approach was that
this use of new technologies can also inspire and enable a new appreciation of linguistic-
structural diversity or pluralism in language use in institutional contexts and thus licence
a new type of language awareness and language policy.

14
 REFERENCE No.
To cite this output:
RES-000-23-1133
Matras, Yaron (2008). Place, mobility, and dialect differentiation in Romani : Full Research Report
ESRC End of Award Report, RES-000-23-1133. Swindon: ESRC
The project’s conceptual orientation was to revisit the relation between speech
form an location, and speech form and community, by allowing a non-hierarchical
representation of a broad dataset representing variation in the language. The presentation
of this dataset to users would raise awareness of variation and diversity, and help set an
agenda for language education and language policy that would embrace pluralism rather
than seek to replace it through imposed or artificial uniformity. At the same time,
interpretation and analysis of the data would provide some key insights into the
processes that are responsible for diversification, these being both language-internal
(processes of analogy, levelling, and grammaticalisation) and language-external (processes
of linguistic convergence and borrowing or ‘fusion’), whose propagation vehicles are the
social networks among speech communities and ultimately among individual speakers as
the most ‘discrete’ loci of language use.

x Objectives
The objectives of the research were
1) To develop theory in the area of historical linguistics and especially contact linguistics,
and to cover the historical, typological, and contact-related development of variation
among the dialects of Romani in publications arising from the project;
2) The develop a web resource for the study of dialect differentiation and to make data
on Romani dialects accessible to a wide audience of users;
3) To further international networking in the study of Romani.

As a means of achieving these goals, the project set out to

4) Carry our fieldwork on dialects of Romani that have not yet been described, focusing
in particular on those that fall in-between what is commonly regarded as ‘prototype’
dialects in traditional model of Romani dialect classification,
5) Engage an international team of fieldworkers and assistants in various countries
6) Evaluate the data, as well as to disseminate relevant parts of the research to a wide
audience of users.

All of these objectives have been met in full:

1) The project has produced a series of publications, among them several book chapters
as well as MA and PhD theses, conference papers, and journal papers, that deal with
contact and convergence theory and with variation within and among the dialects of
Romani (see Results and Outputs below).
2) The project has developed an extensive website with documentation on the Romani
language, which includes text sections on the history of the language, status, dialect
differentiation, and structures, as well as interactive sections with maps and sound
samples and an uptodate bibliographical database, with links to numerous other sites and
acitivities (http://romani.humanities.manchester.ac.uk/). The project’s principal output
is the RMS Database, with a comprehensive documentation, accompanied by examples
in transcription and sound, of Romani samples from over 150 locations across Europe
(http://romani.humanities.manchester.ac.uk/rms). (see Methods and Outputs).
3) The project has contributed to international networking in various ways: by employing
a team of fieldwork assistants in several different countries, by organising workshops and
a cross-disciplinary international conference, and by participating at international events.
(see Activities).
4) Fieldwork on Romani was carried out in several countries (see Methods), making it
possible to extend the project’s data coverage to some 300 samples altogether,
representing over 150 locations in Europe. This is one of the widest ever extents of

15
 REFERENCE No.
To cite this output:
RES-000-23-1133
Matras, Yaron (2008). Place, mobility, and dialect differentiation in Romani : Full Research Report
ESRC End of Award Report, RES-000-23-1133. Swindon: ESRC
coverage of any dialectological project, and the most diverse and logistically challenging,
taking into consideration the need to operate in different countries and in different
contact languages.
5) The international team of fieldworks trained and engaged by the project remains
committed to work on Romani, and individuals within this group have continued to
pursue postgraduate studies with a specialisation in Romani linguistics and cultural
anthropology, while others hold organisational positions in Romani culture initiatives, at
the level of local authorities and non-governmental organisations.
6) A series of publications by the project team and collaborators has emerged in
conjunction with the project and drawing on the resources that the project created. In
part, they form a coherent ‘school’ of work on Romani, addressing issues that arise from
a common discussion platform (see Outputs). But they also invite further contributions
by members of the academic community who have direct access to the project’s
resources and to much of its yet unexploited data corpus.

x Methods

The project’s principal target was to deliver an original, innovative and sophisticated
language documentation tool: The so-called ‘RMS (Romani Morpho-Syntax) Database’
(http://romani.humanities.manchester.ac.uk/rms/). The database figures as a data
archive for well over 300 samples of Romani recorded in part before, and largely during
the lifetime of the project. Each sample represents a speaker from a particular location
and background, and a metadata section covers speaker’s details as well as all relevant
technical and organisational aspects of the particular dataset. The samples represent
varieties of Romani from altogether close to 150 different locations in Europe. Each
sample consists of a list of some 1,080 phrases recorded using a uniform questionnaire,
which is designed to cover all major aspects of morpho-syntactic, phonological and core-
lexical variation in Romani (for the complete questionnaire see
http://romani.humanities.manchester.ac.uk/rms/browse/phrases/phraselist).

For around one third of the samples, short free narratives were also recorded, often
addressing aspects of the speaker’s personal biography or aspects of community customs
– description of marriage, burial, birth and festivities. The phrases are tagged for the
various structural aspects that they represent, e.g. semantic constructions, inflection class,
phoneme inventory, and so on. The tags are programmed to retrieve relevant phrases
into the specific chapters of the database, represented by table cells on specific pages.
These chapters and tables together constitute a comprehensive grammatical description
of each and every sample. It not only provides information on the key relevant structural
features, but also exemplification through the relevant phrases, in transcription,
translation, and sound. The RMS Database is thus a comparative, historical, and
typologically informed reference grammar with examples in transcription and audio, with
unrestricted online access – an unprecedented type of language documentation resource.

The database uses open-source software (MySQL, PHP) and is hosted by a Faculty
server at the University of Manchester. The fact that open-source software was used
makes the code accessible for future modifications and upgrades of the database and
frees it from any dependency on the licensing or version upgrade policy of commercial
software. The database utilises conventional state-of-the-art technology in all aspects, for
example UNICODE (UTF-8) fonts to represent transcription symbols, MP-3 for digital

16
 REFERENCE No.
To cite this output:
RES-000-23-1133
Matras, Yaron (2008). Place, mobility, and dialect differentiation in Romani : Full Research Report
ESRC End of Award Report, RES-000-23-1133. Swindon: ESRC
sound files, and Flowplayer for online video demonstrations of use procedures. The
application is designed to operate on all state-of-the-art web browsers, and the sound can
be played on any of the standard audio applications, thus ensuring that its operation is
system- and platform-independent. The database is accompanied by an extensive help
menu which offers detailed explanations and still image as well as video demonstrations
of all aspects of its use – both for the casual user who is browsing, searching and
querying data, and for the more active participant who is entering, storing and editing
data. It is thus configured to serve both a wide audience of external users and future
generations of ‘insiders’ who will further contribute to using the resource to archive and
edit their own data. The search functions allow users to retrieve any combination of data
from any of the table cells in the database. Search results can be browsed in the normal
table format, as well as downloaded into table in various standard formats (CSV). They
can also be plotted on a map, which, drawing on the Google map system offers a display
of the results of several search queries at a time. This dynamically generated
dialectological map draws on the location coordinates that are assigned automatically to
any specified location in a sample’s metadata set.

The advanced query builder allows users to run queries on larger sets of data and extract
correlations and hierarchies among specified data particulars, which may be defined
themselves by any combination of properties. Thus, the database is able, for instance, to
calculate the probability of borrowing for a particular morpheme, or to identify a
hierarchy in the presence or absence of a particular morpho-phonemic feature in
different points within a paradigm.

In order to prepare the datasets for import into the RMS database, several steps were
undertaken. Fieldwork was commissioned according to the original project plan to target
areas in which there had been gaps in the documentation of Romani dialects. Project
fieldwork assistants were recruited and with the help of contacts established via
specialists (e.g. ethnographers working in Romani communities), NGO activists, or
indeed, in the case of assistants of Romani background, acquaintances and relations in
their own community, speakers were identified in relevant locations who were willing to
cooperate with the project. The following fieldwork assistants were employed in
recording and transcribing data from the following countries: V Schulman (Greece), A
Tenser (Baltic area, Ukraine, Moldova), A Borda (Poland), H Pirttisaari (Finland), G
Beluschi (Italy, Romania), C Raducan (Moldova), F Jacobs (Romania), V Chileva
(Bulgaria), J Jovanivic (Serbia, Montenegro, Croatia), P Baba (Hungary, Ukraine), and a
student team led by V Elšík and consisting of Z Bodnarova, K Frydryskova, B Sebova, E
Vranova, P Rubak, P Zapotocky, M Skocovska, P Kubanik, M Vancakova (Slovakia).

The audio data were delivered in various audio formats, and where necessary were
digitised in Manchester (or usually copies were made in various digital formats, one, in
AIFF or WAV, for storage and another, in MP3, for display on the web). The
transcriptions were delivered in pre-formatted spreadsheets. They were checked carefully
and where necessary revised and edited by a team consisting of the project’s archive
manager and supported by several part-time assistants (H Gardner, Ch Schubert, R Hill,
A Tenser, V Schulman). The sound files were then segmented to match the individual
words and phrases contained in the spreadsheet. The resulting individual files were
uploaded along with each spreadsheet into the database. Each phrase number remained
pre-tagged for the various relevant structural features that it displayed (e.g. semantic type
of clause combination or case relation, time adverbs, deictics, particular inflection classes,

17
 REFERENCE No.
To cite this output:
RES-000-23-1133
Matras, Yaron (2008). Place, mobility, and dialect differentiation in Romani : Full Research Report
ESRC End of Award Report, RES-000-23-1133. Swindon: ESRC
numerals, and so on). The full list of tags can be viewed online:
http://romani.humanities.manchester.ac.uk/db/gramma.html

As mentioned above, each table cell within the database interface was tagged for the
relevant function(s) that it documents. Tags are modifiable and can be assigned by users
with editing rights to individual cells, either by adding the actual tag, or simply by adding
the phrase number that contains a relevant example. Once phrases were uploaded, the
input procedure into the database could begin. Drawing on the tags, the editor/inputter
– the archive manager and part-time support staff – retrieves relevant examples,
identifies the correct form and inserts it into the cell. The result is a cell containing a
grammatical analysis – for example an answer to an analytical question (“Are case
markers retained?”) or the form of a case ending (“-es”) – that is then exemplified by the
accompanying word or phrase, in transcription and sound.

The extremely laborious procedure of fieldwork, data processing, data input and
descriptive analysis was repeated for altogether 92 samples, for which processed data
have been uploaded and inputted (for a complete list see:
http://romani.humanities.manchester.ac.uk/db/dialects.html ). Other samples have been
archived in the project archive; their details can be retrieved via RMS. Alongside this
corpus, RMS contains information gathered from some 60 referenced published sources,
giving altogether representation to some 150 different varieties of Romani.

Through the procedure outlined above, the project achieved two of its principal goals,
namely the documentation and presentation of data on a wide corpus of dialects, and the
descriptive analysis in the form of a typologically informed, historically oriented
descriptive grammar, or this enormous and diverse corpus. Further methods were
applied to the evaluation and the dissemination of data. Several descriptive works
produced by project members and associated students followed the general orientation
and structure of the RMS Database as developed by Elšík & Matras, and the historical
and typological outlook on Romani as represented in the monographs by Matras (2002)
and by Elšík & Matras (2006), and especially by Matras’s grammatical sketch of the
Romacilikanes dialect of Epirus (Matras 2004). Underlying this approach is a certain
understanding and reconstruction of ‘Early Romani’, which determines the blueprint of
developments that are outlined in the description, combined with a commitment toward
a certain extent in the typological coverage of functional categories, and finally a certain
historical understanding of the spread of major isoglosses among Romani dialects, and a
description of individual samples with reference to those.

In addition, a more general functional-typological approach was taken to the evaluation

of cross-dialectal variation, as exemplified largely by Elšík & Matras (2006). This was
followed in a series of papers and in one comparative PhD dissertation. A series of
works on contact, written by the PI, as well as chapters on contact phenomena delivered
by some of the project students, follows an approach to contact that is outlined in the
PI’s recent monograph on Language Contact (currently in press with Cambridge
University Press). Broadly speaking, this approach is anchored in Functional Pragmatics
and an action-oriented understanding of language change, and views so-called structural
hybridity in language as an attempt by speakers to capitalise as far as possible on their full
repertoire of linguistic structures, by partly removing demarcation boundaries among sets
of structures (or ‘linguistic systems’). The focus in applying this approach to the Romani
data was on the predictability of borrowing of certain categories, and the identification of
borrowing hierarchies. From these hierarchies, semantic-pragmatic motivational factors

18
 REFERENCE No.
To cite this output:
RES-000-23-1133
Matras, Yaron (2008). Place, mobility, and dialect differentiation in Romani : Full Research Report
ESRC End of Award Report, RES-000-23-1133. Swindon: ESRC
for borrowing were extracted (as already in work by Matras 1998, 2000, and by Elšík &
Matras 2006).

Finally, a note is in order concerning the method of dissemination of information and

data on Romani to wider audiences. Here, the emphasis was on the use of new
technologies (interactive resources) as well as on the notion of pluralism. The project’s
DVD-ROM (which can be viewed or downloaded on
http://romani.humanities.manchester.ac.uk/files/cd1.shtml ) stresses the processes that
have given rise to variation within the language, exemplifies language use in various
regions of Europe through the display of audio and video samples, and demonstrates
how variation is also found in present-day institutional use of Romani in writing and
broadcasting, conveying the message that pluralism is no hindrance to effective
communication, protection or status-elevation of a smaller language. A similar message is
conveyed in a series of papers by the PI on codification and language policy, and in the
PI’s input into educational and language policy activities (see below).

x Results

The project results are varied and ‘layered’. No doubt the most central contribution of
the project is the creation of the web-based documentation tool, and the descriptive
analysis of some 100 original samples that it contains. Alongside its technological
achievement, the project has delivered results in the academic evaluation of variation in
Romani, and in general raising awareness of Romani and variation within it. Project staff
and associated PhD students have delivered several descriptions of Romani dialects that
have been hitherto undescribed (such as the dialects of Velingrad in Bulgaria, of the Servi
of Ukraine, of the Gabor of Transylvania, and of the Loftika Rom of Lithuania). A
significant breakthrough is the documentation of the Romani dialects of Moldova and of
Italy, on which information was lacking before almost entirely.

A focused analysis on dialects in so-called ‘transition’ zones – the dialects of Ukraine, the
dialects of Vojvodina, and the dialects of Transylvania – served to revise the traditional
notion of dialect ‘branches’ (as represented recently by work by Boretzky & Igla 2004).
Rather than outbranching, the emergence of dialect variation has been interpreted in the
context of the project as the diffusion of individual innovations carried by networks of
social contacts. These networks are dynamic and their shape is ever-changing, which is
the reason why individual isoglosses do not always cluster. For the dialects of Ukraine, it
was shown how speakers entertain a notion of a ‘Ukrainian’ dialect (referred to as ‘Servi’)
while in fact showing variation that combines features that are otherwise seen in the
descriptive literature as diagnostic of separate dialect ‘branches’. A similar result was
identified for the much more specific social-denomination of the ‘Gabor’ Roma or
Transylvania. By contrast, the analysis of the Romani dialects of Vojvodina shows pretty
much a straightforward geographical continuum that bridges between the dialects of
Banat and Serbia and those of Hungary. Taken as a whole the individual regional profiles
confirm that dialects evolve through a network of contacts despite the absence of a
Romani-speaking territorial continuum. Particularly illuminating in this respect is the
comparison between a so-called dialect ‘branch’ – the Northeastern dialects of Poland,
the Baltics, and North Russia – and an individual community in Sofades, Karditsa,
Greece. Variation is found within the community, correlating in part with speakers’
external contacts. As for the ‘branch’, it displays some core shared features, alongside a

19
 REFERENCE No.
To cite this output:
RES-000-23-1133
Matras, Yaron (2008). Place, mobility, and dialect differentiation in Romani : Full Research Report
ESRC End of Award Report, RES-000-23-1133. Swindon: ESRC
tendency for each location to share some features with neighbouring regions outside the
Northeastern cluster itself, thus introducing variation into the cluster. As a result, the
very notion of ‘dialect’ was given up by the project as a rather constructed entity, giving
way to an objective investigation of the relations among individual ‘samples’, as personal
forms of speech. Romani dialectology is thus understood as the evaluation of the
structural profiles of such individual samples and their correlation with a series of
language-external attributed such as location, migration history, customs and self-
identification.

Comparative investigations have traced the emergence and divergence of categories

marking modality and transitivity, and special attention has been paid to the predictability
of contact influence and implications for a general theory of contact, and beyond, toward
a general theory of cultural hybridity. Contact hierarchies were determined through a
comparison of the contact behaviour of categories within the sample, and these were
interpreted in terms of the semantic-pragmatic advantages for the speaker in allowing
‘fusion’ of the structural representation of categories across languages. Typological issues
that have received attention in the evaluation include the representation of modality and
sources for the grammaticalisation of modals as well as borrowing of modals, clause
combining and the representation of semantic relations of factuality as well as borrowing
susceptibility of connectors, and transitive morphology and the semantic maps that its
development follows. For the latter, the concept of ‘intensity’ was introduced for verbs,
with its concrete manifestations ranging from causativity through to iterativity and target-
oriented action (telicity), and on to the adaptation of loanverbs (as an explicit allocation
of the attribute of ‘verbness’).

A series of publications highlighted the relevance of linguistic pluralism to language

policy in Romani and urged the incorporation of a concept of pluralism into language
education activities.

x Activities
The project relied during its work on a network of collaborators based in different
countries, who collected and processed data. It organised a number of training
workshops and hosted an international cross-disciplinary conference on Romani studies
in Manchester in September 2007. Project members participated at the Romani studies
conference in Granada in September 2005 and at the 7th International Conference on
Romani Linguistics in Prague in September 2006. The project invited leading
ethnographers E. Marushiakova and V. Popov to consult on the design and method of
elicitation of narratives describing community customs, and for discussions on the
evaluation of the material in regard to a model of consolidation and segregation of
Romani communities.
The project’s launch event for its website in January 2006 was attended by the
Lord Mayor of Manchester and by representatives of the Romani communities and local
authorities, and received wide press coverage (see
http://romani.humanities.manchester.ac.uk/files/61_launch_press.shtml). Radio reports
on the project’s activities followed in Australia, Hungary, US, Austria, and on a series of
national and regional programmes in the UK, including Rokker Radio (Romani radio
programme), Five Live, and various programmes of BBC Radio 4. BBC World television
broadcast a feature on the project in February 2006. Additional media coverage appeared
following the launch of the project’s DVD in May 2007. Project members participated in

20
 REFERENCE No.
To cite this output:
RES-000-23-1133
Matras, Yaron (2008). Place, mobility, and dialect differentiation in Romani : Full Research Report
ESRC End of Award Report, RES-000-23-1133. Swindon: ESRC
a series of community events, including the Roma/Gypsy History Month and in
consultations at local schools with a population of Romani pupils. The PI was invited to
give numerous seminar presentations concerned with the project, in Budapest, Prague,
Paris, Bergamo, Eisenstadt, Istanbul, Helsinki, Granada, Hamburg, Leipzig, Cluj, and
Melbourne, and was invited to consult the Committee for the Charter of Minority and
Regional Languages of the Council of Europe as well as the European Roma and
Traveller Forum (permanent representation of Roma at the Council of Europe), in
November 2005, and the Council of Europe’s Language Policy Division.

x Outputs
The project’s major electronic outputs are the RMS database, the Romani Linguistics
Webpage, and the interactive DVD-ROM on the Romani language. Written outputs
include 3 MA theses, 3 PhD dissertations, 1 published grammatical description, and a
number of conference presentations and book chapters (see attached Annex). A major
journal article examining geographical diffusion of structural innovations in Romani
dialects is still in preparation and envisaged for submission by the end of November
(awaiting feedback at a presentation at the International Conference on Romani
Linguistics in St Petersburg in September 2008). A plan to submit for publication a book
on the dialects of Romani was withdrawn; the book would have largely duplicated the
material already available via the online database. Instead, a further journal article is
planned that will outline an historical reconstruction scenario of the processes that led to
dialect differentiation in Romani.

x Impacts

The project cooperated closely with the Council of Europe and the Romani
representation there, as well as with the Committee for the European Charter of
Minority and Regional Languages, and was invited to draft language policy documents
for Romani. These documents highlight the importance of linguistic pluralism in
addressing language codification and the promotion of Romani language in the education
system. The PI’s input into the discussions helped inspire the creation of a Council of
Europe working group, which authored a Language Curriculum Framework for Romani.
The justification paper for the Framework was drafted by the PI. This framework, which
embraces the concept of pluralism in language, is currently being adopted by several
education authorities around Europe.

In cooperation with the UK Department for Education and Skills and a number of
NGOs and international foundations, the project created the first ever multilingual
educational resource – a DVD-ROM in 17 different languages entitled ‘The Romani
language: An interactive journey’. The DVD was widely disseminated internationally and
is being used within the education system in several countries. Some 3,500 printed copies
were distributed directly by the project to schools, NGOs and interested individuals, and
some 1,500 copies were downloaded from the project’s website, around half of those in
the full multilingual version, the other half in individual language versions, with Romani,
Russian, and Spanish in the lead. The DVD is not copy-protected and it can be assumed
that further dissemination by users has taken place.

The project’s website has been serving as the principal source of information on Romani
worldwide, and has had some 45,000 unique visits from over 70 different countries since

21
 REFERENCE No.
To cite this output:
RES-000-23-1133
Matras, Yaron (2008). Place, mobility, and dialect differentiation in Romani : Full Research Report
ESRC End of Award Report, RES-000-23-1133. Swindon: ESRC
its launch in 2006. Since the RMS-Database went online on the site in January 2008,
traffic has increased to an average of 200 visits per day, representing a daily average of
some 3,500 hits (i.e. attempts to access data forms or pages on the site). Since January
2008, over 60% of hits – or a total of over 750,000 – have come from outside the UK,
confirming the project’s international standing as the principal resource on the Romani
language.

The project has also had an impact on training in Romani linguistics. Altogether 3 MA
dissertations and 3 PhD dissertations were completed as part of the project. Two of the
MA students are of Romani origin (O Chashchikhina from Russia, V Chileva from
Bulgaria), the first Romani women in the world to complete an MA degree with
specialisation in Romani linguistics. The project also boosted interest in the
Undergraduate course unit on Romani linguistics at the University of Manchester, which
saw record enrolments of a total of 270 students during the project’s lifetime.

x Future Research Priorities

Despite the enormous coverage by the RMS Database enabled through this project, there
remain gaps in the documentation of Romani dialects, most notably in the dialects of
France, Belgium, northern Italy, and Germany (the so-called Sinti cluster), and in the
description of the dialects of Turkey, as well as Albania and the southern Peloponese. It
is hoped that a future project will be able to close these gaps, drawing on the existing
fieldwork and documentation tools developed during this project. In addition, the
database facility can still be expanded to allow the incorporation of discourse excerpts
(narratives or so-called ‘texts’). Much of the material exists already in the project’s
archive, and the technological solution to allow incorporation of this material is there as
well. What is needed is staff time in order to process and tag the data. This will allow in
the future not just to broaden the basis for exemplification of structural phenomena, but
also to devote more time to the cross-dialectal study of discourse-level phenomena such
as deixis, discourser markers, and word order variation. Finally, still unexplored in depth
is the historical relationship between Romani and Domari. The study of the Romani and
Domari dialects of Anatolia, where the two languages meet, may provide a key to this
question.

22
 REFERENCE No.
To cite this output:
RES-000-23-1133
Matras, Yaron (2008). Place, mobility, and dialect differentiation in Romani : Full Research Report
ESRC End of Award Report, RES-000-23-1133. Swindon: ESRC
Annex: List of project outputs

Matras, Yaron
2005. The status of Romani in Europe. Report submitted to the Council of Europe’s
Language Policy Division, October 2005.
2005. Language contact, language endangerment, and the role of the ‘salvation linguist’
In: Austin, Peter K. ed. Language documentation and description, Volume 3. London:
Hans Rausing Endangered Languages Project. 225-251.
2005. The future of Romani: Toward a policy of linguistic pluralism. Roma Rights
Quarterly 1:31-44.
2005. The classification of Romani dialects: A geographic-historical perspective. In:
Halwachs, D. & Schrammel, Barbara, Ambrosch, Gerd. eds. General and applied
Romani linguistics. Munich: Lincom Europa. 7-26.
2006. with Viktor Elšík. Markedness and language change: The Romani sample. Berlin:
Mouton de Gruyter. 475pp.
2006 entries on Romani and Domari. In: The Encyclopedia of Language and Linguistics.
Second edition. Oxford: Elsevier.
2007. The challenges of language codification in a transnational context. In: Molinelli,
Piera, with Giuliano Bernini, Pierluigi Cuzzolin, Ada Valentini. Eds. Standard e non
standard tra scelta e norma. Atti del XXX congresso della Società Italiana di Glottologia.
Roma: Il Calamo. 43-54.
2007. Contact, connectivity and language evolution. In: Rehbein, Jochen, Hohenstein,
Christiane & Pietsch, Lukas. eds. Connectivity in Grammar and Discourse. Amsterdam:
John Benjamins. 51-74.
2007. The borrowability of grammatical categories. In: Matras, Y. & Sakel, J. eds.
Grammatical borrowing in cross-linguistic perspective. 31-74.
2007. with Jeanette Sakel. Investigating the mechanisms of pattern replication in
language convergence. Studies in Language 31, 829-865.
2007 with Gilad Margalit. Gypsies in Germany – German Gypsies? Identity and
politics of Sinti and Roma in Germany. In: Stauber, Roni & Vago, Raphael, eds.
The Roma: A minority in Europe. Historical, social and cultural perspectives. Oxford:
Berghahn. 61-82.

2008. with Jeanette Sakel. Modelling contact-induced change in grammar. In: Stolz,
Thomas, Bakker, Dik, Salas Palomo, Rosa. eds. Aspects of language contact. Berlin:
Mouton de Gruyter. 63-87.
in press A contrast language? Re-examining linguistic hybridity in Romani. In: Elšîk,
Viktor. ed. Proceedings of the 7th International Conference on Romani Linguistics. Munich:
Lincom.
in press. Universals of structural borrowing. In: Siemund, Peter. ed. Linguistic universals and
language variation. Berlin: Mouton.
in press. Contact, convergence and typology. In: Hickey, Raymond. ed. Handbook of
Language Contact. Oxford: Blackwell.
in press with Viktor Elšîk. Modality in Romani. For: Hansen, B., de Haan, F. & van der
Auwera, J. eds. Modality in European languages. Berlin: Mouton.
in press with Christopher White and Viktor Elšîk. The RMS Database and web resource.
In: Everaert, Martin & Musgrave, Simon. eds. Linguistic databases. Berlin: Mouton
de Gruyter.

Yaron Matras: Relevant seminars and conference presentations:

23
 REFERENCE No.
To cite this output:
RES-000-23-1133
Matras, Yaron (2008). Place, mobility, and dialect differentiation in Romani : Full Research Report
ESRC End of Award Report, RES-000-23-1133. Swindon: ESRC
2005 Invited keynote speaker: Language codification in a transnational context: the
case of Romani. Annual meeting of the Societa Italiana di Glottologia, Bergamo,
20-22 October 2005.
2005 Grammatical borrowing in Domari. Conference on Grammatical borrowing in
universal perspective. University of Manchester 30 September - 1 October 2005.
2005 ‘Migration, Diasporas, and Role of Language as Memory: The Case of the Roma
(Gypsies)’. Workshop on ‘Narrative, memory, visibility’. School of Languages,
Linguistics and Cultures. University of Manchester. 25 May 2005.
2005 Endangered languages, contact, and the ‘better linguist’: The case of Domari.
Workshop on Endangered Languages, Hans Rausing Endangered Language
Documentation Project, SOAS, February 2005.
2006 Mixed Languages: structural and functional constraints. Naples University, June
2006.
2006 Methods in the dialectology of Romani. Molise University, Isernia, June 2006.

2006 The origin of the Rom, Lom and Dom. International Symposium on Roma in
Turkey. Istanbul, 6-7 May 2006.
2006 Language policy in a transnational context: The case of Romani. Workshop on
Regionalism, Nationalism, Transnationalism. University of Manchester, 10-11
March 2006.
2006 A contrast language? Toward a model of linguistic hybridity in Romani, 7th
International Conference on Romani Linguistics, Charles University, Prague. 15-
17 September 2006.
2007 Toward an integrated approach to grammatical borrowing. RCLT Seminar, La
Trobe University, November 2007
2007 Documenting the Romani dialects of Europe. Institute for Advanced Study, La
Trobe University, Melbourne, June 2007
2007 Language revitalisation and the documentation of Angloromani. York University
Linguistics Seminar, May 2007
2007 The language of the Gabor: Implications for Romani dialectology. International
conference on ‘Romani diasporas, Romani migrations’, University of Manchester,
September 2007
2007 Variation, bilingualism and contact-induced change. Workshop on Variation and
Morphosyntactic Change in Contact Situations, Paris (Research Ministry),
September 2007
2007 Clause combining in Epirus Romani. International workshop on ‘Clause
Linkage’, Research Centre for Linguistic Typology, La Trobe University,
Melbourne. August 2007.
2008 Defining ‘Everyday multilingualism’. Keynote address at the opening event of the
UNESCO Year of Languages, Eisenstadt, Austria, June 2008
2008 Documenting the Romani varieties of Britain and Europe. Language in Context
seminar, University of Edinburgh, May 2008
2008 Language contact: and integrated theory. Hebrew University of Jerusalem
Linguistics Seminar, March 2008
2008 A functional approach to language contact. Linguistics Seminar, Charles
University, Prague, March 2008

Tenser, Anton
2005 Lithuanian Romani. Munich: Lincom
2006 Romani dialects in Ukraine and Moldova : reconciling linguistic and

24
 REFERENCE No.
To cite this output:
RES-000-23-1133
Matras, Yaron (2008). Place, mobility, and dialect differentiation in Romani : Full Research Report
ESRC End of Award Report, RES-000-23-1133. Swindon: ESRC
ethnographic data. 7th International Conference on Romani Linguistics,
Charles University, Prague. 15-17 September 2006.
2008 The Northeastern dialects of Romani. PhD dissertation, University of
Manchester

Schrammel, Barbara
2005. The classification of Romani dialects: A geographic-historical perspective. In:
Halwachs, D. & Schrammel, Barbara, Ambrosch, Gerd. eds. General and applied
Romani linguistics. Munich: Lincom Europa. 7-26.
2006 From valency to aktionsart : the marker -ker- in East Slovak Romani. 7th
International Conference on Romani Linguistics, Charles University, Prague. 15-
17 September 2006.

Schulman, Veronica
2006 The system of loan verb markers (LVM) in Sofades Romani (Greece). 7th
International Conference on Romani Linguistics, Charles University, Prague. 15-
17 September 2006.
2008 The Romani dialect of Sofades (Karditsa, Greece). Unpublished PhD
Dissertation, University of Manchester

Beluschi Fabeni, Giuseppe

2006 Roma Korturare, kaj žanas le verdonenca : some ethnographical answers to the
Romani dialectological survey. 7th International Conference on Romani
Linguistics, Charles University, Prague. 15-17 September 2006.

Chileva, Veliyana
2005 The Romani dialect of Velingrad (Bulgaria). Unpublished MA thesis, University
of Manchester

Chashchikhina, Olga
2006 The Servi Romani dialects of Ukraine. Unpublished MA thesis, University of
Manchester

White, Christopher
2008 Database development document (2nd version). Ms., University of Manchester

Brown, Stuart; Gardner, Hazel; Jones, Charlotte; Matras,Yaron

2007 The Romani language: An interactive journey. DVD-ROM (Flash). University of
Manchester.

25