Eriksson_etal_maniscript.

advertisement
A comparison of the use of relativization in the Swedish regional
dialects spoken in Burträsk and Ström
Anders Eriksson,
Department of Linguistics, Gothenburg University, Sweden,
Fredrik Karlsson and Kirk P. H. Sullivan,
Department of Philosophy and Linguistics, Umeå University, Sweden.
Mailing address
Anders Eriksson
Department of Linguistics
Gothenburg University
Box 200
SE 405 30 Gothenburg
Sweden
Phone: +46 31 773 5232
E-mail: anders.eriksson@ling.gu.se
Short title: The use of relativization in two Swedish dialects
The use of relativization in two Swedish dialects
1
A comparison of the use of relativization in the Swedish regional
dialects spoken in Burträsk and Ström
Abstract
In this paper we compare the usage of a common type of relativizer by speakers of two Swedish
dialects. The data are drawn from a large database of recorded speech from 107 Swedish
dialects. The speakers represent two generations and an equal number of male and female
speaker were represented in each generation group. It was expected that there would only be
minor differences between the two dialects and that gender and age differences would follow the
traditional lines with female speakers being less dialectal than male speakers and young speakers
less dialectal than older ones. Both expectations turned out to be completely wrong, however.
The two dialects showed significant differences with respect to the usage of the studied
relativizer in all comparisons whereas intra-dialectal variation as a function of gender and age
revealed no significant differences in any of the comparisons.
Acknowledgement
We thank the Swedia 2000 research project for granting us full access to the Swedia 2000
database of dialect recordings. We also want to thank Tomas Landgren who transcribed the
Ström recordings.
1
The use of relativization in two Swedish dialects
1
2
Introduction
The data presented in this paper come from tape recordings made as part of the dialect project
“The phonetics and phonology of the Swedish dialects around the year 2000, SWEDIA 2000”,
funded for a six-year period by The Bank of Sweden Tercentenary Foundation. Karlsson and
Sullivan (2002) have investigated relative marker usage in Burträsk in a previous study. Due to
the location of Burträsk at the northern periphery of the Swedish dialect rectangle and as a noncoastal settlement, they believed that the Burträsk dialect may well reveal traces of older variants
of Swedish. As the SWEDIA dialect database is balanced for informant gender and age, it
permitted the investigation of both the infiltration of standard Swedish relativizers into this
northern dialect of Swedish and the influence of gender and age in any other change that may
have occurred during the twentieth century. Karlsson and Sullivan posed the following two
specific research questions:
1. Is the use of relativizer in the Burträsk dialect dependent upon the gender of the speaker?
2. Has there occurred a change in relativizer construction and usage in the Burträsk dialect
over the past fifty years?
From the tape-recordings of eight informants’ spontaneous speech lasting a total of 6 hours and
55 minutes, Karlsson and Sullivan found a strong preference in the Burträsk dialect for relative
clause constructions involving a moving (realized) subject, that relativization involving object
movement is also used, but less frequently, and that in relative clauses involving the null relative
marker, subject and object correlates occur with equal frequency. Further, no significant
difference in relative marker usage was found due to either age or gender. Of particular note was
that no examples of the use of constructions using a wh-word, such as vars or vilken/vilket were
2
The use of relativization in two Swedish dialects
3
found in the data provided by any of the eight informants. Unlike, the complementizer som that
was found in the data vars or vilken/vilket always imply a more formal, or standard, speaking
style.
This paper represents a substantial extension of the previous study. We begin by posing the same
questions in relation to the dialect spoken in Ström as we did for the Burträsk dialect. The focus
in this study, however, is on contrasting the Ström data with the Burträsk data with respect to the
usage of relative constructions by the speakers of these two dialects. Like Burträsk, Ström is a
non-coastal settlement. It lies around 350 km southwest of Burträsk and has since the middle
ages been a trading centre. Of particular interest is that Ström lies in Jämtland, which was part of
Norway until 1645. The comparative data presented in this paper may provide a picture of how
the use of relativization is affected when a settlement has contact with neighbouring dialects due
to its position as a trading centre and due to its less peripheral location within the Swedish
Dialect Rectangle
This paper begins by presenting the SWEDIA dialect database, before over-viewing
relativization in Standard Swedish. Thereafter, the syntactic variation that can be found in the
Swedish dialects is illustrated by a few examples from the Skellefteå dialect before we present a
brief description of the two communities where the studied dialects are spoken, technical details
about the recordings and biographical details about the informants. Then the method of data
analysis is presented and illustrated by examples from both dialects and finally, the Burträsk and
Ström data are compared and contrasted.
3
The use of relativization in two Swedish dialects
2
4
SWEDIA 2000
The Swedia 2000 project is a national research project investigating the phonetics and phonology
of Swedish dialects. It is a joint venture shared between the departments of phonetics at Lund,
Stockholm and Umeå universities. The database from which the material in this study was drawn
consists of recorded material from 107 dialects evenly distributed over Sweden and the Swedish
speaking parts of Finland recorded during 1998–9. Twelve speakers were recorded for each
dialect. The speakers were selected so as to reflect a relevant variation in age and gender. The
age dimension was meant to reflect variation between two generations of speakers - younger
speakers, and older speakers roughly thought of as being the parent generation of the younger
speakers. The ‘young’ speakers were defined as speakers in the age range 25 to 35 years of age.
The older speakers were (with few exceptions) in the age range from 55 to 70 years of age. An
equal number of speakers (3) in each category were recorded. To guarantee the representativity
of the recordings from a dialect point of view the informants must have been born in the area and
lived there for the major part of their lives. For the younger speakers it was a requirement that
their parents should also have been born and lived in the area for most of their lives. This
criterion was not strictly applied for the older speakers, but was nevertheless fulfilled for most of
them.
The speech material consisted of a controlled part – word lists designed to elicit material from
which phonological systems and stress systems may be reconstructed for the different dialects –
which was the same for all dialects and a ‘spontaneous speech’ part which consisted of informal
interviews or conversations between two informants.
4
The use of relativization in two Swedish dialects
5
All the tape-recordings were made in the informants’ homes by experienced research
assistants working for the project. The research assistants ensured that the tape-recordings were
made in situations that, as far as was practically possible, were free from background noise and
interference from other speakers.
3
Relativization
Relative clause construction in the Germanic languages has been studied from a range of
different viewpoints as is clearly illustrated in Poussa (2002) and the varying approaches
represented by the following studies. Engdahl (1997) studied the formal definitions of the
mechanisms of correlate extractions, Kayne (1994) and Platzack (1997) both examined the
structure of restrictive and non-restrictive relative clauses, and Poussa (2001) and Van den
Eynden Morpeth (2001) both presented diachronic studies of the use of relative markers.
Taken as a whole this body of research demonstrates that speakers of different Germanic
languages behave differently with respect to relative construction in many ways. This fact when
coupled with the diffuse distinction between what separates language vs. what separates regional
dialects makes contrastive study of the use of syntactic units in different regions speaking the
same language, as well as their distribution in a diachronic perspective, worthy of investigation.
An example of such a study is the one presented in Poussa (2001) on the English dialect spoken
in Docking, north-west Norfolk. The present study contrastively studies the use of relative
markers in the Swedish dialects spoken in Burträsk and Ström, and examines the distribution of
5
The use of relativization in two Swedish dialects
relative markers in these dialects diachronically through comparison of the older and younger
speakers of these dialects around the year 2000.
3.1
Relativization in Standard Swedish
Standard Swedish uses two distinct types of realized relative markers: wh-words such as vars in
(1) and vilken in (2) and a complementizer, som, as in (3). Both the wh-words, vars and vilken,
and the complementizer, som, can be used as both restrictive and non-restrictive relatives
(Platzack (1997, 2002). Tables 1 and 2 (Platzack 2002), respectively, show the cross-Germanic
syntactic variation for restrictive and non-restrictive relatives, and the options available in
Standard Swedish.
(1)
Mannen,
vars
bil
var
The man, whose car
gul,
gick ut
was yellow, went out
med hunden.
with the dog.
`The man, whose car was yellow, went out with the dog.’
(2)
Mannen
vilken
jag
pratade med var
The man
who
I
talked
väldigt trevlig.
with was very
nice.
`The man who I talked to was very nice.’
(3)
Hunden
som mannen
ägde
var
svart.
The dog
that the man
owned was black.
`The dog that the man owned was black.’
(INSERT TABLE 1 ABOUT HERE)
6
6
The use of relativization in two Swedish dialects
7
(INSERT TABLE 2 ABOUT HERE)
Apart from the realized relative markers, Swedish permits the use of a  (null) relative
marker in object relativization as indicated in Table 1 and example (4).
(4)
Det var
It
bollen
was the ball

hon sparkade.
REL she
kicked.
`It was the ball {that/} she kicked.’
It should be pointed out at this point that use of the null relative marker was also permitted in
subject relativization in Old Swedish as can be seen in example (5) (Platzack, 2002). The
corresponding sentence is not grammatical in modern Standard Swedish (6).
(5)
Alla the mæn [thæta breff høra ælla see]
All
the men this
letter hear or
see
`All the that hear or see this letter’
(6)
*Alla de män [detta
brev hör eller ser]
All
letter hear or
the men this
see
‘All the that hear or see this letter.’
The use of a d-pronoun as a relative marker, as used in some of the Germanic languages, is
disallowed in Standard Swedish. As can be seen from Tables 1 and 2, Swedish is similar to
English rather than, for example, Dutch in the use, or lack of use, of the d-pronoun. In Dutch, die
7
The use of relativization in two Swedish dialects
8
is the d-pronoun used to introduce all relative clauses whose heads are not neuter singular, hetwords. The choice of d-pronoun depends upon the number and gender of the head of the relative
clause. This is illustrated in (7), (8) and (9) using examples taken from Karlsson and Sullivan
(2002).
(7)
Boken
som
jag
läser
är väldigt
intressant.
Het boek
dat
ik
lees
is geweldig
interessant.
The book <NEUTER,SINGULAR> I
am reading is very
interesting.
`The book that I am reading is very interesting’.
(8)
Böckerna
som
jag
läser
är
De boeken
dat
ik
lees
zijn geweldig
The books
<NEUTER,PLURAL> I
am reading are
väldigt
very
intressant.
interessant.
interesting.
`The books that I am reading are very interesting’.
(9)
Tidningen
som
jag
läser
är väldigt
intressant.
De krant
dat
ik
lees
is geweldig
interessant.
The newspaper
<NEUTER,SINGULAR> I
am reading is very
interesting.
`The newspaper that I am reading is very interesting’.
3.2
Swedish regional dialects
In common with other languages, the many regional dialects of Swedish differ from Standard
Swedish phonetically, phonologically, morphologically and syntactically. That is, in addition to a
8
The use of relativization in two Swedish dialects
9
subset of the properties of standard Swedish that each dialect contains in order to be classified as
Swedish, it also contains a range of features that gives the dialect its unique identity. Karlsson
and Sullivan (2002) exemplified these using sentences in the dialect of Skellefeå. We use the
same examples here. In example (10), a difference in phonetic realization between Standard
Swedish and the dialect of Skellefteå, the different derivational morpheme for the verb used in
this dialect and a difference in the inflectional morpheme of the object noun phrase between
Standard Swedish and the Skellefteå dialect are exemplified. The third row in (10) gives an
English word-by-word translation of the dialect sentence.
In addition to the above-mentioned differences, there are also vocabulary differences between
Standard and regional dialect Swedish. This is exemplified in (11), again comparing the
Skellefteå dialect with Standard Swedish and taken from Karlsson and Sullivan (2002).
(10)     
 
 +e
+n
(Skellefteå
dialect)
DET.MASC.   
 +   (Standard
+
Swedish)
DET.MASC peter
kick
+ed
ball
the.MASC
‘Peter kicked the ball.’
(11)        
(Skellefteå dialect)
        (Standard Swedish)
I
put
in the nets a little
further away.
‘I put the nets in a little further away.’
9
The use of relativization in two Swedish dialects 10
4
Burträsk and Ström
The community of Burträsk is situated in the north of Sweden, 90 km north of Umeå and 40 km
south-west of Skellefteå (figure 1). Surrounding the community itself are a number of smaller
communities where people live, but seldom work or go to school. The children of the area go to
school in Burträsk for school years 1—9 and in Skellefteå or Umeå for the non-compulsory
school years 10—12.
INSERT FIGURE 1 ABOUT HERE!
Figure 1. A map showing the dialect area Burträsk. The small map of the Nordic countries in the
upper right corner shows the location of the Burträsk area.
The community of Ström is situated in northern Jämtland and was part of Norway until 1645.
Ström lies approximately 350 km south-west of Burträsk (see Figures 1 and 2) and is situated in
the most northern of Jämtland’s municipal districts, Strömsund that was formed from the
merging of Frostviken, Fjällsjö, Ström and Hammerdal municipal districts. As in Burträsk, there
are a number of smaller communities surrounding Ström where people live and attend school
during school years 1 - 9. After the nine compulsory years of schooling, all the students in the
municipality of Strömsund attending school for years 10 - 12 attend Hjalmar Strömerskolan in
Strömsund. As of December 31, 2000, 5 677 people lived in the modern parish of Ström and a
total of 13 938 people lived in the municipality of Strömsund.
10
The use of relativization in two Swedish dialects 11
INSERT FIGURE 2 ABOUT HERE!
Figure 2. A map showing the dialect area Ström. The small map of the Nordic countries in the
upper right corner shows the location of the Ström area.
5
The tape-recordings
The tape-recordings used in this study were the Burträsk and Ström dialect recordings from the
SWEDIA 2000 tape-recording archive. The part of the recordings used in the present study was
what in paragraph 3 was mentioned as the ‘spontaneous speech’ part.
Relevant technical and informant data are summarised in Table 3. Eight (out of twelve)
informants for each dialect - four male and four female, and four from the older subgroup and
four from the younger subgroup - were selected for the present study. The number of sentences is
to be seen as approximate since the occurrence of hesitations and re-starts makes the decision
whether a new sentence has begun, or not, a subjective one. The only reason for not including all
twelve speakers in the study was lack of personal resources to do so. A more detailed description
of each informant is given in Section 5.1.
INSERT TABLE 3 ABOUT HERE!
Six of the eight Burträsk informants were interviewed by a SWEDIA 2000 research assistants
who, although they did not speak the Burträsk dialect, spoke a generic northern dialect of
Swedish and were capable of understanding the informants without difficulty. The two
remaining subjects (ym1 and ym2) were recorded together in a dialogue style session. The
research assistant only participated in the discussion by asking questions when the dialogue
11
The use of relativization in two Swedish dialects 12
between the two informants began to flag. The same conditions applied in the Ström recordings.
Here informants yw2 and ym2 were recorded talking to each other rather than being interviewed.
5.1
The informants
Tape recordings of eight informants in each of the two dialects were analysed. Below follows a
summary of relevant background information about the participating informants.
Burträsk
5.1.1
YM1
YM1 was born in Ljusvattnet in 1975. His father came from Ljusvattet, outside Burträsk, and his
mother came from Norsjö, a small community 50 kilometres north-west of Ljusvattnet. YM1
attended school years 1–9 in Burträsk and then went on to study the High School programme in
construction for four years in Skellefteå, a larger town 30 kilometres north-east of Ljusvattnet.
He then moved back to Burträsk and worked as a craft, design and technology teacher until he
began his current job as a carpenter for a woodworking company.
When asked about his use of dialect in day-to-day situations he stated that he would normally
use his dialect when speaking to others except when it was obvious that the person listening did
not understand what was being said.
Apart from his native dialect, YM1 considered himself to be competent in standard Swedish
and English.
5.1.2
YM2
YM2 was born and raised in Gurkoträsk (25 km west of Burträsk) in 1975. His father was born
and raised in Gurkoträsk and his mother was born in Storås in Jämtland. After finishing school
years 1–9 in Burträsk he moved to Skellefteå where he attended High School for three years and
12
The use of relativization in two Swedish dialects 13
lived for a further 1.5 years after finishing High School. He then moved back to Burträsk to take
up his current job in a sawmill.
Subject YM2 stated that he uses his native dialect in most situations, except when going to
places where he is not understood (e.g. the Swedish capital Stockholm). Apart from his native
dialect and standard Swedish, YM2 reported moderate proficiency in English.
5.1.3
YW1
YW1 was born in 1978 in Bygdträskliden, a small community 10 km south-west of Burträsk.
Her father was born in Brännbergsliden and her mother in Innansjön. After nine years of
schooling in Burträsk, she moved to Skellefteå where she continued her studies for three more
years, after which she moved back to Burträsk. Here, before taking up her current job as a health
care worker, she worked customer relations.
She reported that she used her native dialect in all situations, and that she also has some
proficiency in English.
5.1.4
YW2
YW2 (born 1968) grew up in Burträsk where she attended school for the nine years of
compulsory education. Her father was born in Floda and raised in Burträsk and her mother was
born and raised in Rentjärn outside Malå (110 km north-west of Burträsk).
YW2 lived in Stockholm for eight months when she was 20 and was living in Nedre Åbyn at the
time of recording.
After completing her two-year High School course (school years 10–11), she worked first as a
postal worker and then a health care worker, and at the time of the recordings was a student at
Komvux (a type of University Access Course). Apart from her native dialect, which she
13
The use of relativization in two Swedish dialects 14
reported using in all day-to-day situations, she viewed herself to be competent in standard
Swedish and English.
5.1.5
OM1
OM1 was born in Ljusvattnet in 1926 where, apart from a three year period when he lived in
Klemensnäs, he had lived all his life. Both his parents were from Ljusvattnet and after finishing
seven years of schooling, he started working in Ljusvattnet as a sawmill worker. Later he worked
as a carpenter until he retired.
He stated that he never used anything else than his native dialect.
5.1.6
OM2
OM2 was born in 1943 in Ljusvattnet where he also spent most of his youth. His father was born
in Ljusvattnet and his mother came from Ljusheden. After finishing his seven years of
schooling, OM2 spent one year doing his national service in Umeå. He then found a job as a
quality inspector’s assistant for six months in Trelleborg, after which he moved back to Burträsk
and bought a farm in Södra Åbyn.
He stated that he used his native dialect in most situations, except formal ones and on trips to
larger cities such as Umeå and Stockholm where he believed he uses language that is more
similar to Standard Swedish than he would normally use.
Apart from his native dialect and some Standard Swedish, OM2 also claimed to have some
proficiency in English.
5.1.7
OW1
OW1 was born in Renbergsvattnet in 1924 where she lived until 24 years of age, when she got
married and moved to Ljusvattnet. Her father was also from Renbergsvattnet and her mother
14
The use of relativization in two Swedish dialects 15
was from Åbyn. After finishing the mandatory seven years of schooling, she worked first as a
health-care worker and later as a day-carer and as an artist.
She stated that she spoke her native dialect in all situations expect those when the people she
is speaking to would not understand if she used her dialect.
5.1.8
OW2
OW2 was born in 1935 in Häggnäs where she spent most of her youth. Her mother was from the
same village and her father was from Brännvattnet. She trained as a health-care worker and
remained in the profession until retirement.
She stated that she used her native dialect on all occasions except when talking to medical
doctors.
Ström
5.1.9
YM1
YM1 was born in 1971 in Strömsund, where he lived until the age of six when the family moved
to Falun in Dalarna for a period of four years, after which they moved to Hillsand (33 km
northwest of Strömsund), where both his father and mother had their roots.
His educational background included the compulsory school years 1-9, and High School (years
10-12) where he specialized in Economics and trained as a nursing assistant. YM1 was living in
Hillsand and was working as a professional carer at the time of recording.
When asked about his use of his regional dialect, YM1 reported that he would use his dialect in
conversation with friends and family speaking the same dialect, but not, for instance, in
telephone conversations and when talking to people he did not know or when visiting another
15
The use of relativization in two Swedish dialects 16
area of Sweden. He viewed himself as being somewhat proficient in speaking English and
German.
5.1.10
YM2
YM2 was born in Strömsund in 1966. His father was from the smaller community of Hillsand
outside Strömsund and his mother was originally from Vilhelmina. When YM2 was seven years
old, the family moved from Strömsund to Hillsand. After school years 1-9, YM2 went on to
High School in Strömsund in order to train as a builder. After High School, YM2 received
additional training as a toolmaker. YM2 was living in Hillesand at the time of recording. YM2
stated that he used his regional dialect when talking to friends speaking the same dialect as well
as when talking to members of the family. Apart from his regional dialect and standard Swedish,
he regarded himself to be fairly proficient in English.
5.1.11
YW1
YW1 (born 1974) lived in Strömsund until the age of 15, when she moved to Östersund. Her
mother was from Strömsund and her father from Svaningen (55km northwest of Strömsund). She
completed all her schooling (years 1-12) in Strömsund. During the final three years she trained
to work with horses. She was living in Strömsund and working in a cafe when the recording was
made. YW1 stated that she used her regional variant of Swedish mostly within the family and
that she did not use it at work.
5.1.12
YW2
YW2 was born in 1975 in Ulriksfors (2.5km southeast of Strömsund) where she spent most of
her youth. Her mother was from Offerdal and her father was from Jalasjärvi in Finland.
16
The use of relativization in two Swedish dialects 17
Her schooling background consisted of the nine years of mandatory education and three years of
Senior High School followed by supplementary courses in computer management. She was
working as a computer consultant and living in Strömsund at the time of recording.
She reported being proficient in English and moderately proficient in French. She stated that she
used her regional dialect at all times, but that she would modify her speech slightly at work for
the sake of clarity.
5.1.13
OM1
OM1 was born in 1943 in Renålandet, where both his father and mother were from.
Prior to moving to Strömsund five years before the recording, he had lived for all but one year of
his life in Renålandet. The year he spent was away from Renålandet was 1963 which he spent in
Gaza as a part of a UN mission in 1963. In his youth, he received 7 years of schooling and then
trained as a builder. He reported being moderately proficient in English and that he used his
regional dialect in everyday situations, but that he would use a more standard variant of Swedish
when people did not understand him.
5.1.14
OM2
OM2 was born in 1941 in Strömsund and apart from16 months spent on a UN mission, he had
lived there his entire life. His father was from Siljåsen (40km north of Strömsund) and his
mother from Strömsund. He had received seven years of schooling and training as a rescue
worker. He worked as lumberjack and truck driver for most of his life. OM2 stated that he would
use his regional dialect and that he was moderately proficient in English.
5.1.15
OW1
OW1 was born in 1924 in Äspnäs (18km northwest of Strömsund), where she spent most of her
youth. Her mother was from Äspnäs and her father from Laxsjö, Fölinge (39 km west of
17
The use of relativization in two Swedish dialects 18
Strömsund). Her educational background consisted of seven years of schooling and her training
in health care management. She worked within the health care system all her professional life.
She reported using her regional dialect in all situations, but that she would use a variant of
Swedish that is more like standard Swedish when in situations when her dialect was not easily
understood. Apart from her regional dialect and standard Swedish, she reported knowing words
and short phrases in English.
5.1.16
OW2
OW2 was born in 1952 in Strömsund, where she had lived for most of her life, although at the
time of the recording she was living in Svaningen. She has also lived for four months in Luleå
and four months in Sundsvall. Her father was from Strömsund and her mother was from
Svaningen. Her education background consisted of the compulsory school years 1-9, two years
of High School and health care training. She stated that she used her regional dialect when
talking to friends and family, but that she would use a more standardized variant of Swedish
when the visiting areas in Sweden where she would not be understood.
5.2
Data analysis
The eight tape-recordings for each dialect were transcribed orthographically and all the relative
clauses were extracted from the transcriptions and analysed. For the Burtäsk material, one of the
orthographic transcriptions (ow1) was then re-examined by an independent researcher (a
Swedish syntactician) and the level of agreement between the two analyses was investigated in
order to ascertain the standard error in the analysis. Inter-researcher agreement was 100% for
subject relativizers, 50% for object relativizers and for 33% for NULL relativizers. All of the
inter-researcher disagreement for the NULL relativizers consisted of a relative clause with a
NULL relativizer being assigned by the first researcher and not the second researcher. The
18
The use of relativization in two Swedish dialects 19
sentences for which there was inter-researcher disagreement were extracted from the
orthographic transcription. These, along with the two analyses, were then presented to the
independent researcher who was asked to make a judgement as to which of the two alternative
analyses was the correct one. The extracted sentences were presented randomly, as were the
order of the two alternative analyses. This meant that neither researchers’ analysis was always
presented first. Hence, no selection bias was introduced. The data were presented as a `two
alternative, forced choice’ (2AFC) test. The resulting analysis made by the independent
researcher was then judged to be the template analysis as the differences between the two
analyses were resolved after a brief discussion between the two researchers. The entire Burträsk
dataset was then reanalysed according to the template analysis and 100% agreement between the
analysis and the template was achieved.
Regarding the Ström material, initial transcriptions and categorisation of relative markers were
made by a native speaker of the Ström dialect. The transcriptions and categorisations produced
by the transcriber were then re-examined by the first and second author. Those differences in
judgments that were found, were discussed and a final analysis was produced in which all the
differences had been resolved.
6
6.1
Use of relativization Burträsk and Ström
Burträsk
he informants in the Burträsk dialect recordings used only one explicit relativizer. These
constructions used the complementizer som as the relative marker. Examples of relative clauses
with realized relative markers are illustrated in examples (12–15). Interestingly, no examples of
19
The use of relativization in two Swedish dialects 20
the use of wh-words (vars, vilken) as the relative marker were found in the tape-recordings; this
option is present in Standard Swedish and is discussed in Platzack (2002).

   (om1)
(12)  

It
was
.
a
government
department
that
was
here.
‘It was a government department that was here.’
(13)
  
  


   (ym1)


 .

There are indeed many who only move themselves in this
village.
‘There are indeed many who only move within this village.’
   
(14) j 


 (ym2)
 

Yes we have a couple of real
originals who
live up in Åbyn.
‘Yes, we have a couple of really original people who live in Åbyn.’
(15)  
 
  
   (ow1)

 
 
.
  
20
The use of relativization in two Swedish dialects 21
But the vinter clothing that then the men had had in the forest.
‘But the winter clothing that the men had had then in the forest.’
The Burträsk recordings did show usage of the unrealized (Null) relative marker. The unrealized
(Null) relative marker was used in both subject constructions (example 16) and object
constructions (example 17).




 
(16) 
We have a
pretty
cool
vicar
 .
(ym1)
REL plays
golf.
`We have a pretty cool vicar who plays golf.’


 
(17) 
Yes it
was leather goods






. (ow1)
REL she
made
for a
company
in town.
‘Yes, it was leather goods that she made for a company in town.’
6.2
Ström
As was the case with the Burträsk dialect speakers, the Ström dialect speakers only used one type
of explicit relativizer, the complementizer som. Examples of som used as a subject relativizer is
shown in 18–19 and as an object relativizer in 20–21.
(18)   
In
a
village
that


(ow1)
is-called
21
Äspnäs
The use of relativization in two Swedish dialects 22
In a village that is called Äspnäs
(19)    

it

 
is indeed not
all
 
(ym2)
 
who
travel to Västerås
It is indeed not everyone who travels to Västerås

 

 

(20) 
That was
indeed an
episode
that
I
(ow1)
remember
That was indeed an episode that I remember

 





(21) 
(ym2)
This here
is indeed a
system
that
I
not
am used to in fact
This is indeed a system that I am not used to
In contrast to what seems to be the case with the Burträsk dialect, unrealized relativizers only
occured in object constructions in the Ström dialect recording. Two such cases are exemplified in
22–23.

(22) 

this here
station
REL



 (om2)
I
work
this station (that) I work at now
22
at
now
The use of relativization in two Swedish dialects 23
(23) 



it
is
a

  



one-of-those REL
you

should chug
(yw2)

around on
It’s one of those you should (just) chug around on
6.3
Quantitative results and statistical analyses
As may be seen in Table 3, the total number of sentences, produced by each speaker varies
considerably. To make comparisons between speakers and dialects possible, this variation has to
be eliminated by some kind of normalization procedure. This was achieved by using the relative
occurrence of relativizers, defined as the proportion of sentences (in percent), which contained a
relativizer for each speaker as the quantitative measure. These data are summarized in Table 4.
INSERT TABLE 4 ABOUT HERE!
In the following analyses only dialect (Burträsk - Ström), speaker category (male - female,
young - old) and type of relativizer will be considered. In the underlying analyses, however, the
full set of data has always been used. In all analyses, “sentences containing a relativizer” will be
taken to mean also those sentences where the relativizer receives a null-realization on the
surface.
A first analysis of the data in Table 2, showing the proportion of relativizers by dialect and
speaker category is presented in Figure 3.
INSERT FIGURE 3 ABOUT HERE!
It is immediately clear from Table 2 and Figure 3 that all four categories of speakers of the Ström
dialect use relatively more relativizers than the Burträsk speakers. A full factorial univariate
analysis of variance using dialect, age, and gender as independent variables, shows the apparent
23
The use of relativization in two Swedish dialects 24
dialect difference to be highly statistically significant (P = 0.011). There are, however, no
significant contributions by the two other factors (age, gender), nor are any of the interactions
between the factors statistically significant. As may be seen in the diagram, there are minor intra
dialect differences with respect to the extent to which relativizers are used by the male/female,
younger/older speakers, but none of these differences reaches statistical significance.
INSERT FIGURE 4 ABOUT HERE!
Figure 4 shows the relative occurrence of relativizers broken down by type. It may be seen that
the realized subject relativizer is by far the most frequently used type, but also that the Ström
speakers use relativizers not only more frequently in general but that this is also the case for
three of the four different types of relativizers. A MANOVA using the relative occurrence of the
four relativizer types as dependent variables and dialect, age and gender as independent variables
shows that the differences in the frequency of occurrence is statistically significant for all four
types as a function of ‘dialect’ (P < 0.05 in all comparisons) with the Ström dialect speakers
producing more relativizers of all kinds except null-subject relativizers where the Burträsk
dialect speakers produce more. As in the previous analysis, the factor age does not make a
significant contribution in any of the comparisons (P = .24 or greater). The same is true for
gender with respect to both types of object relativizers and for the realized subject relativizer (P
= .23 or greater). For null-realizations of subject relativizers, gender makes a significant
contribution (P = 0.048). It must be pointed out, however, that this difference is an indirect effect
caused by the fact that in Ström there is only one occurrence of this type of relativizer (om1 uses
24
The use of relativization in two Swedish dialects 25
it once) whereas in Burträsk five of the eight speakers use this type at least once. The difference
should thus be attributed to the variable dialect rather than gender.
An intra dialect analysis reveals that Burträsk speakers use realized subject relativizers
significantly more than any of the other three types but that there is no significant difference in
usage between those types. For the Ström dialect speakers the differences in the frequency of
usage is statistically significant for all four types. Age and gender make no significant
contribution in any of the intra dialect analyses.
INSERT FIGURE 5 ABOUT HERE!
The diagram shown in Figure 5 is essentially a summary of what is already shown in Figure 4,
but here the main categories, subject and object relativizers, are pooled. The differences between
dialects are significant for both types (P = .001 and .042 respectively). No other factors
contribute significantly.
INSERT FIGURE 6 ABOUT HERE!
Figure 6 presents the same data but from a different perspective. The data underlying this
diagram are the means of the proportions (in percent) to which each speaker uses the four kinds
of relativizers. Now a different picture from the one shown in Figure 5 emerges. Whereas the
Ström dialect speakers use both subject and object relativizers significantly more often, they use
a proportionally smaller number of subject relativizers and a proportionally greater number of
25
The use of relativization in two Swedish dialects 26
object relativizers than do the Burträsk dialect speakers. Again the differences with respect to
dialect are significant (P = .026 and .027 respectively), whereas age and gender do not contribute
significantly.
7
Discussion and conclusion
The aim of the present study was to compare the use of relativizers in two Swedish dialects. The
presentation above has focused primarily on the differences between the two studied dialects, but
to put the differences into perspective we must acknowledge that there are, of course, also many
basic similarities. Speakers of both dialects only use the complementizer som in relative
constructions. No examples of the use of constructions using a wh-word, such as vars or
vilken/vilket were found in the data provided by any of the eight informants. This is in contrast to
the situation found in Standard Swedish (as well as its neighbouring Germanic languages
German, Dutch, English, Icelandic, Faeroese, Danish and Norwegian) where occurrences of a
wh-word as a relative marker have been observed.
Interestingly, native speakers of Swedish feel that constructions using a wh-word, such as
vars, mark a more formal (and standard) speaking style. Indeed, as Platzack (2002) wrote: “The
use of a wh-word, even the possessive version, always implies formal style”(our emphasis).
This could explain why no examples of wh-word relativization were found in the recordings
analyzed here. The complementizer som does not, per se, imply a more formal, or standard,
speaking style and thus. The complementizer som is, according to Platzack (2001), “the
unmarked choice”.
26
The use of relativization in two Swedish dialects 27
Both dialects use realised relativizers more than unrealised ones and subject relativizers are
more common than object relativizers.
But given these general similarities, there are also many quite striking, and rather unexpected,
differences. When we study the use of relativizers in closer detail, the dialects turn out to differ
in all comparisons made. The Ström dialect speakers use relativisers to a greater extent. They
also use three of the four the different types more frequently. One type, however, the unrealised
subject relativizer, common in Old Swedish but no longer used in most dialects is still used in the
Burträsk dialect, albeit not frequently, but is hardly found at all in Ström (only a single
occurrence). Although both dialects use subjects relativizers more frequently than objects
relativisers, the proportions of subject vs. object relativizers used differ significantly between the
two dialects.
The most surprising finding, however, is perhaps the fact that none of the differences between
the two dialects may be attributed to the age or gender of the speakers. Nor has the study
revealed any intra dialect differences with respect to age or gender. The conclusion that may be
drawn from this finding is that the use of relativisers is extremely stable in both dialects and no
change over time is apparent at least from the data analysed here. Given the rapid change in
dialects in many other respects this finding is quite surprising. An explanation one may propose
is the following:
As the informants come from within a single socio-economic group and are educated to
similar levels, the lack of difference due to gender could be interpreted as implying that relative
markers are not an indication of social status within the dialect-speaking population. If this were
the case, it might be expected, following the work of Eckert (1989), Labov (1990), Holmes
(1997), Docherty et al. (1997) and Nevalainen (2000), that the female population tended more
27
The use of relativization in two Swedish dialects 28
towards the socially accepted standard more than the men but no evidence to support any claim
of this sort was found here. This lack of social importance of the relative marker as a status
marker within the dialect community, is possibly reflected in lack of change in the pattern of
frequency and distribution of use of the relative markers between the older and the younger subgroups. This finding can be interpreted as indicating that there have been no major changes in
relative marker use over the past 50–70 years in the two dialects studied here.
To what extent these findings are representative of Swedish dialects in general is not possible
to say. Very little research has do date been done on grammatical properties of Swedish dialects.
Too little is therefore known about grammatical similarities and differences with respects to
Swedish dialects, not just concerning relativizers but grammar in general. In the introduction we
raised the question whether differences, if any, found in the Ström dialect when compared with
the Burträsk dialect could be interpreted as a reflection of Ström’s contacts with neighbouring
dialects due to its role as a trading centre or possibly a result of Norwegian influence. As has
been demonstrated in the present study, the Ström dialect differs significantly from the Burträsk
dialect in all aspects where the two dialects were compared. So the first part of the question, is
the Ström dialect any different from other dialects, say the one spoken in Burträsk, has been
answered in the affirmative. To answer the second part of the question, can the observed
differences be attributed to Norwegian influence or influence from other neighbouring dialects
caused by Ström’s position as a trading centre remains to be investigated. In a wider perspective
one would want to be able to place the differences and similarities found in Burträsk and Ström
in the context of a typology including many more Swedish dialects, but much more research is
needed to be able to do that. The investigation presented here is, however, a step in that
direction.
28
The use of relativization in two Swedish dialects 29
29
The use of relativization in two Swedish dialects 30
8
References
Docherty, Gerard J., Foulkes, Paul, Milroy, James, Milroy, Lesley. & Walshaw, David. (1997).
Descriptive adequacy in phonology: a variationist’s perspective. Journal of Linguistics
33:275–310.
Eckert, Penelope. (1989). The whole woman: sex and gender differences in variation. Language
Variation and Change 1:245–267.
Engdahl, Elisabet. (1997). Relative clause extractions in context. Working papers in
Scandinavian Syntax 60:51–79.
Holmes, Janet. (1997). Setting new standards: sound changes and gender in New Zealand
English. English World-Wide 1:107–142.
Karlsson, Fredrik & Sullivan, Kirk P. H. (2002). The use of relativization in the regional dialect
of Swedish spoken in Burträsk. In P. Poussa (Ed.), Relativisation on the North Sea
Littoral. Munich, Germany: Linkom Europa. 77–107.
Kayne, Richard S. (1994). The antisymmetry of syntax. Cambridge, Mass: MIT Press.
Labov, William. (1990). The Intersection of sex and social class in the course of lingusitic
change. Language Variation and Change 2:205–254.
Nevalainen, Terttu. (2000). Gender differences in the evolution of Standard English. Journal of
English Linguistics 28:38–59.
Platzack, Christer. (1997). A representational account of restrictive and non-restrictive relatives:
the case of Swedish. Working papers in Scandinavian Syntax 59:65–97.
30
The use of relativization in two Swedish dialects 31
Platzack, Christer. (2002). Relativization in the Germanic languages, with a particular emphasis
on Scandinavian. In P. Poussa (Ed.), Relativisation on the North Sea Littoral. Munich,
Germany: Linkom Europa. 77–96.
Poussa, Patricia. (2001). Syntactic change in north-west Norfolk. In J. Fisiak, J. & P. Trudgill
(Eds.), East Anglian English. Cambridge, U.K.: D. S. Brewer. 243–259.
Poussa, Patricia. (Ed.) (2002). Relativisation on the North Sea Littoral. Munich, Germany:
Linkom Europa.
Van den Eynden Morpeth, Nadine. (2002). Relativisers in the southwest of England—with
special emphasis on subject contact clauses, preposition placement with particle and
pronominal relative strategies and pushdown relatives. In P. Poussa (Ed.), Relativisation on
the North Sea Littoral. Munich, Germany: Linkom Europa. 181–194.
31
The use of relativization in two Swedish dialects 32
Tables
Table 1. Restrictive Relatives: Cross-Germanic Syntactic Variation (after Platzack, 2002) NSR=
non-subject relativizer, SR=subject relativizer. A shaded square indicates that no such relativizer
is permitted.
Relative
marker
d-pronoun
wh-word
cpl
null
Germ.
Dutch
der, die,
die, dat
das, dessen
welcher
wiens,
wie
Engl.
Icel.
Faroese Dan.
whose (hvers) hvörs
who
that
sem, er sum, ið
NSR
NSR/SR
32
Norw.
der
(der)
hvilken,
hvis
som, at
NSR
(hvilken
hvis)
som
NSR
Swed.
vars,
vilken
som
NSR
The use of relativization in two Swedish dialects 33
Table 2. Non-restrictive Relatives: Cross-Germanic Syntactic Variation (after Plazack 2002)
NSR= non-subject relativizer, SR=subject relativizer. A shaded square indicates that no such
relativizer is permitted.
Relative
marker
d-pronoun
wh-word
cpl
Germ.
Dutch
der, die,
die, dat
das, dessen
welcher
wiens,
wie
Engl.
Icel.
Faroese Dan.
der
whose (hvers) hvörs
who
sem
sum
null
33
Norw.
Swed.
(der)
hvilken, (hvilken vars,
hvis
hvis)
vilken
som
som
som
The use of relativization in two Swedish dialects 34
Table 3. Age, recording length in minutes and number of sentences for each subject analysed.
The subjects ym1 and ym2 in Burträsk and yw2 and ym2 in Ström were recorded together in a
dialogue setting. This is indicated by the star (*) in the time-column.
Informants
Age
BURTRÄSK
Recording
Number of
length
Sentences
Old woman 1 (ow1)
74
0:37
297
Old woman 2 (ow2)
65
0:36
160
Old man 1 (om1)
72
0:51
310
Old man 2 (om2)
55
0:31
317
Young woman 1 (yw1)
21
0:59
332
Young woman 2 (yw2)
30
0:39
148
Young man 1 (ym1)
24
2:42*
304
Young man 2 (ym2)
23
2:42*
312
6:55
2180
Total
STRÖM
Old woman 1 (ow1)
75
0:38
370
Old woman 2 (ow2)
47
0:44
477
Old man 1 (om1)
56
0:31
298
Old man 2 (om2)
58
1:03
677
Young woman 1 (yw1)
25
0:38
422
Young woman 2 (yw2)
24
1:06*
472
Young man 1 (ym1)
28
0:43
362
Young man 2 (ym2)
33
1:06*
270
5:23
3348
Total
34
The use of relativization in two Swedish dialects 35
Table 4. The table shows the proportion of sentences (in percent) containing a given type of
relative marker used by each informant. Each realized or unrealized token is categorised
according to the function of the referenced word. All of the realized relative markers are som.
Speaker
Subject relativzers
Object relativizers
Realized (%)
Null (%)
Realized (%)
Null (%)
Total (%)
ym1
6.9
0.3
0.0
0.3
7.6
ym2
8.7
1.0
0.6
1.0
11.2
yw1
4.2
0.0
0.3
0.3
4.8
yw2
2.7
0.0
1.4
0.7
4.7
om1
3.5
0.3
0.0
0.3
4.2
om2
7.9
0.3
0.3
0.0
8.5
ow1
4.0
0.3
0.7
0.3
5.4
ow2
3.8
0.0
0.0
0.0
3.8
5.21
0.28
0.41
0.37
6.27
Burträsk
Mean
Ström
ym1
12.7
0.0
2.5
1.1
16.3
ym2
5.2
0.0
1.1
0.4
6.7
yw1
3.1
0.0
1.4
0.9
5.5
yw2
6.8
0.0
2.3
1.9
11.0
om1
8.7
0.3
3.0
0.7
12.8
om2
8.7
0.0
1.8
1.3
11.8
ow1
10.3
0.0
2.4
1.4
14.1
ow2
12.8
0.0
0.8
0.4
14.0
0.04
1.93
1.01
11.51
Mean
8.53
35
The use of relativization in two Swedish dialects 36
Figure 1
Figure 1. A map showing the dialect area Burträsk. The small map of the Nordic countries in the
upper right corner shows the location of the Burträsk area.
36
The use of relativization in two Swedish dialects 37
Figure 2
Title:
Creator:
ArcView Version 2.1
Preview :
This EPS picture w as not saved
w ith a preview included in it.
Comment:
This EPS picture w ill print to a
PostScript printer, but not to
other ty pes of printers .
Figure 2. A map showing the dialect area Ström. The small map of the Nordic countries in the
upper right corner shows the location of the Ström area.
37
The use of relativization in two Swedish dialects 38
Figure 3.
Percent sentences that contain relativizers
16
14
12
10
Ström
Burträsk
8
6
4
2
0
ym
yw
om
ow
Figure 3. The average proportion of sentences (in percent) containing a relativizer, by dialect
and speaker category.
38
The use of relativization in two Swedish dialects 39
Figure 4.
Frequency of occurrence of the four types
9
8
7
6
5
Ström
4
Burträsk
3
2
1
0
S Real
S Null
O Real
O Null
Figure 4. Proportion of all sentences (in percent) that contain a given type of relativizer, by
dialect and type. (S Real = realized subject realativizers, S Null = non-realized subject
realativizers, O Real = realized object realativizers, O Null = non-realized object realativizers).
39
The use of relativization in two Swedish dialects 40
Figure 5.
Frequency of occurrence of the two
grammatical categories of relativizers
10
8
Ström
Burträsk
6
4
2
0
Subject
Object
Figure 5. Relative occurrence of relativizers, expressed as the proportion (in percent) of all
sentences that contain a given type, by main category, subject vs. object.
40
The use of relativization in two Swedish dialects 41
Figure 6.
Proportional use of the two grammatical
categories of relativizers
100
90
80
70
60
50
40
30
20
10
0
Ström
Burträsk
Subject
Object
Figure 6. Proportional use of the two main categories of relativizers. The bars represent the
means per dialect of the proportions used by each speaker.
41
Download