Nordic Atlas of Language Structures (NALS) Journal,
Vol. 1
Licensed under a Creative Commons
Attribution 3.0 License
Empirical data resources
1. Introduction
We encourage all authors to use the Nordic Dialect Corpus and the Nordic
Syntax Database. These resources were developed over many years (2005-2012)
under the Scandinavian Dialect Syntax (ScanDiaSyn)
and NORMS (Nordic Centre of Excellence in Microcomparative
Syntax) networks of researchers. The research councils in the individual
countries funded most of the recordings, and the technical development of both
data resources, by the Text Laboratory, UiO, was funded by the
Research Council of Norway under the Norwegian Dialect Syntax (NorDiaSyn) project. All researchers, regardless of
nationality or affiliation, are entitled to use these two resources. Access is given by a simple online
registration page at: http://omilia.uio.no/swamp/index.php. The homepage gives a vast amount
of information on the huge amount of effort performed by individuals in the
Nordic countries in the many phases of the creation of these resources:
recordings, transcription, grammatical tagging, technical
solution. It also explains how to search the corpus and database. See: http://www.tekstlab.uio.no/nota/scandiasyn/
1.1 Motivation
the ScanDiasyn project and empirical
resources
Traditional dialectology of the last century focussed
on the study of phonetics, phonology, morphology and lexicon. Although several
monographs on Nordic dialect syntax (e.g. Ragnvald Iversen’s book on the syntax of the Tromsø dialect in 1918, Martin Svendsen’s book on the syntax of the Stavanger dialect in 1931, and Gudrun Lundström’s book on
the syntax of Nyland in
1939) were written at the beginning of the 20th century, syntax was largely
neglected. The ScanDiaSyn network originated from an understanding
among both dialectologists and syntacticians that syntactic variation deserved
more comprehensive investigation. At the same time, while dialectological
research has traditionally been carried out within national borders, one of the
goals of the ScanDiaSyn network has been to study the
dialects across all the Nordic, North Germanic languages: Icelandic, Faroese,
Norwegian, Danish, and Swedish. The latter three languages are mutually
intelligible and their dialects form a continuum. Faroese and Icelandic retain,
to different degrees, many of the characteristics of older stages of the other
languages. Thus, it has proved advantageous to study varieties of all the five
languages under a common approach.
Important outcomes of the ScanDiaSyn collaboration are the two language
resource databases that have been developed specifically for linguistic
research. The Nordic Dialect Corpus (NDC)
(Johannessen et al. 2009) consists of recordings of spontaneous speech from all the
Nordic countries, altogether nearly 2.8 million words, by 797 informants in 223
places across five Nordic countries. It is available on the web through a
user-friendly interface. The Nordic
Syntax Database (NSD) (Lindstad et al. 2009) consists of evaluations of
sentences by 888 informants in 198 places; each informant has evaluated on
average 150 sentences presented in their dialect, grading them on a scale from
1 (unacceptable) to 5 (well-formed). The linguistic contents of both resources
have come about as a result of national subprojects. The Text Laboratory at the
University of Oslo has been responsible for the technical development, and both
resources are available here: http://www.tekstlab.uio.no/nota/scandiasyn/. Section 3 below provides some more
details on these two resources.
The Nordic Atlas of Language
Structures (NALS)
consists of articles on a wide array of syntactic phenomena, and each of them have their empirical base in the Nordic Syntax Database,
with additional evidence from the Nordic Dialect Corpus. The concept of the
NALS web site is similar to that of the The World Atlas of Language Structures (WALS) in that we present maps and commentaries for different linguistic
phenomena. The maps have been generated
from the NDC and the NSD using Google Maps.
2. About the Nordic Syntax
Database and the Nordic Dialect Corpus
2.1. The Nordic Syntax Database (NSD)
The Nordic Syntax Database (NSD) consists of judgments by 812 Nordic
dialect speakers of the Norwegian, Swedish, Danish, Faroese, Icelandic and Övdalian languages. The speakers, who come from altogether
192 different places, have judged a list of sentences designed to test a
variety of different syntactic phenomena. Many of the speakers have contributed
to both the database and the corpus (see below). The sentences in the database
have been assigned grades on a numerical scale, and based on the scores,
dialect maps can be generated, and isoglosses drawn. The judgments can be
sorted and filtered in many ways according to place, age, sex of informants or
type of syntactic phenomenon. The grading scale goes from 1 to 5, where 1
represents an informant’s rejection of a sentence, while 5 represents an
informant’s acceptance of a sentence. In
general, the maps generated show the average evaluation of all informants for
each place. Thus if two informants judged a sentence as 1 and two judged it as
5, the average is 3. The default view of the maps is to show only those places
where the average is 4-5 (i.e. judged as acceptable) and those where the
average is 1-2 (i.e. judged as unacceptable). Sometimes this covers up
interesting variation such as differences across age groups of across sexes.
The NALS authors are aware of this and some papers also discuss such non-geographical
variation.
2.2. The Nordic Dialect Corpus (NDC)
The Nordic Dialect Corpus (NDC) consists of spontaneous speech data from
dialects of the North Germanic languages across all the Nordic countries. The
linguistic data in the corpus come from a variety of sources, both old and new,
but a significant part of the material has been collected as part of the ScanDiaSyn collaboration. The corpus contains nearly 2.7
million words from conversations and interviews by dialect speakers. The
material is transcribed and linked to audio and video recordings, and corpus
has a map function, and it can be searched in a variety of different ways. Even
though the corpus has been developed specifically to facilitate Nordic syntax
research, it is a multipurpose corpus and can be used in a wide range of
research areas, such as phonology, morphology and lexicography. Whereas the NSD
provides information about the same given constructions across at least one
country and often several countries, the NDC has spontaneous speech where the
speakers talk freely either with another informant or with a fieldworker.
2.3. Informants
A number of informants have both evaluated sentences for the Nordic
Syntax Database and been recorded in the Nordic Dialect Corpus. The Norwegian
data have for the most part been collected during the project NorDiaSyn, with the aim of having four informants in each
place (two young and two old). Each informant took part in recorded
conversations and grammatical questionnaires during the same session. The
Danish data were collected under the DanDiaSyn
umbrella, and again all informants (up to six older people) produced both kinds of data. The
Swedish speech data (from both sexes and all age groups) were collected by the
phonology project Swedia2000 and generously handed over to the projects under
the ScanDiaSyn umbrella. The recordings have been
transcribed by the SweDiaSyn project, and as part of
the NorDiaCorp project many of the informants from
Swedia2000 were contacted again about ten years later
and provided grammaticality judgements for the syntax database. Some Icelandic
recordings were provided by Háskóli Íslands, and the project IceDiaSyn
also contributed questionnaire data. The Faroese recordings were done as part
of a fieldwork organised by the NORMS project in
2008. Additional recordings were conducted during similar fieldwork trips to Senja (Northern Norway), Western Jutland (Denmark), and Älvdalen (Sweden). Furthermore, some older Norwegian
recordings from the dialect archive at the University of Oslo (Målførearkivet) have also been included in the corpus with
transcriptions fincanced by the project Norsk Ordbok 2014.
Informant codes used in both the database and the corpus have a
descriptive format, but they differ somewhat according to how and when the data
were recorded. Here are some examples: place, informant-number, age group and
gender. Age is divided into young (under
30) and old (over 50). In both NDC and NSD basic biographical information is
provided for each informant.
|
aal_01um |
Informant no. 1
from Ål (Norway), young man |
[u=ung ‘young’, m=mann ‘man’] |
|
aamot_ma_01 |
Informant no.1
from Åmot (Norway), old recording |
[ma=målførearkivet – the old dialect archive in Oslo] |
|
aarhus3 |
Informant no. 3
from Århus (Denmark) |
|
|
aasen_48 |
Informant no. 48
from Åsen, Älvdalen (Sweden) |
|
|
aaseral_02uk |
Informant no. 2
from Åseral (Norway), young woman |
[u=ung ‘young’, k=’kvinne ‘woman’] |
|
ballangen_03gm |
Informant no. 3
from Ballangen (Norway), old man |
[g=gammel ‘old’,
m=mann ’man’] |
|
bornholm4 |
Informant no. 4
from Bornholm (Denmark) |
|
|
botnhamn_03 |
Informant no. 3
from Botnhamn, Senja (Norway) |
|
|
faro_ym1 |
Informant no. 1
from Fårø (Sweden), young man |
[y=young, m=man] |
|
floby_om1 |
Informant no. 1
from Floby (Sweden), old man |
[o=old, m=man] |
|
fuglafjoerdur_f12 |
Informant no. f12
from Fuglafjörður (Faroe Islands) |
|
|
harbooere_46 |
Informant no. 46
from Harboøre, Western Jutland (Denmark) |
|
|
heroeyMR_03gm |
Informant no. 3
from Herøy, Møre and Romsdal (Norway), old man |
|
|
iceland_a1 |
Informant no. a1
from Iceland |
|
|
iceland_b1 |
Informant no. b1
from Iceland |
|
|
klaksvik_k7 |
Informant no. k7
from Klaksvík (Faroe Islands) |
[k=kvinne ’woman’] |
|
north-iceland_01um |
Informant no. 1
from North Iceland, young man |
|
|
reykjavik_06uk |
Informant no. 6
from Reykjavik (Iceland), young woman |
|
|
st_anna_ow2 |
Informant no. 2
from St.Anna (Sweden), old woman |
[o=old, w=woman] |
References
Johannessen, Janne Bondi, Joel
Priestley, Kristin Hagen, Tor Anders Åfarli, and Øystein Alexander Vangsnes.
2009. The Nordic Dialect Corpus - an Advanced
Research Tool. In Jokinen, Kristiina
and Eckhard Bick (eds.): Proceedings of the 17th Nordic Conference of Computational Linguistics
NODALIDA 2009. NEALT Proceedings Series Volume 4.
Lindstad, Arne Martinus; Nøklestad, Anders;
Johannessen, Janne Bondi; Vangsnes, Øystein Alexander. 2009. The Nordic Dialect Database:
Mapping Microsyntactic Variation in the Scandinavian
Languages. In Jokinen, Kristiina
and Eckhard Bick (eds.): Proceedings of the 17th Nordic Conference of Computational Linguistics
NODALIDA 2009. NEALT Proceedings Series;Volume
4.
Web sites
Nordic Dialect Corpus: http://www.tekstlab.uio.no/nota/scandiasyn/index.html
Nordic Syntax Database: http://www.tekstlab.uio.no/nota/scandiasyn/index.html