12.03.2014
The Norwegian experience The Norwegian experience (& ideas) • 6 municipalities • 1480 pupils at 13 schools • 280 teachers • 52 grown-ups at meetings/clubs etc.
Ragnhild Lie Anderson & Edit Bugge & Helge Sandøy • In total 1812 respondents University of Bergen • (The Åsane-experiment included, 180 resp., cf. Copenhagen 24.2.2014 Anderson in Danske talesprog 10)
1 2
The six localities (pairs: rural-urban)
• Midsund – Molde
• Øygarden – Bergen
• Hå - Stavanger
3 4
Varieties of interest
Set of voices Test A: Rural -old Regionalcentre- A1 Midsund & Molde High (urban) National centre / B1 Molde SE Rural - young Regional centre - SE Low (urban) A2 Øygarden & Bergen B2 Bergen SE A3 Hå & Stavanger B3 Stavanger SE Test B: Regional centre - National centre / High (urban) SE - High Rural Regional centre - National centre / Low (urban) SE - Low • The SE voices are the same
5 6
1 12.03.2014
Test procedure Test voices Questionnaire 1: Topic: invitation - on the mobile phone - to go to the cinema 1. VGT / SEE 2. «What do you think this test was about?» Filled in on an extra (only female voices) page Questionnaire 2: 15 sec voice + 15 sec pause etc. .... LRT: Information: This test was about dialect / language • The varieties distributed as ABCDE CAEBD ADCBE 3. Consciously offered attitude: «Please rate the dialects in the following table based on your opinion of how most people in Norway would evaluate their prestige . ‘Prestige’ refers to • A selection from hundreds of recordings. The selection whether a dialect has a high or low social value. The value of 1 of voices and cuts were tested in our project group in represents the highest score, and 8 represents the lowest order to avoid «test noice» (other possibly influecing score.» factors than the variety) 4. Consciously offered attitude: « How nice do you think the dialects mentioned below are? On the scale, 1 corresponds to
7 ‘nicest’, and 8 corresponds to ‘least nice’.» 8
LRT: Consciously offered attitudes (nice)(blue=Southeastern Norway, results Norway=Oslo region, red=regional centre, green:rural)
Øygarden Bergen (N = Hå Stavanger (N = Midsund • As the Danish ones? (N = 117) 215) (N = 42) 87) (N = 66) 1 Øygarden: Bergen: 2,27 Hå: 2,52 Stavanger: 2,33 Midsund: 1,34 Yes, as far as the consciously offered attitudes are 1,60 i.s 2 Bergensk: Stavanger: Frogner: 2,95 Molde: 2,73 concerned Fana: 2,75 1,70 3,17
3 Austlandsk: Frogner: Sandnes: 3,30 Eiganes: 3,56 Sunnmøre: 2,71 3,69 2,95 4 Arna: 3,93 Frogner: 3,56 Drammen: 3,57 Austlandet: 2,98 5 Sotra: 4,18 Eiganes: 4,10 Sandnes: 3,64
Drammen: Hå: 4,94 9 Drammen: 4.19 4,36
Consciously offered attitudes (prestige/status)(blue=Southeastern SEE: The two dimensions Norway=Oslo region, red=regional centre, green:rural) (dynamism-superiority)
Øygarden Bergen Hå Stavanger Midsund Molde (N = (N = 113) (N = (N = 41) (N = 83 ) (N =65 ) 178) 222) • So far no Danish pattern 1 Bergen: Frogner: Frogner: Frogner: Austland: Aust: 1,76 (but still some work to do with most of the test sites) 1,59 2,04 2,22 1,68 1,65 2 Austland: Bergen: Stavanger: Drammen: Sunnmøre: Trøndelag: 1,92 2,94 2,85 2,88 2,41 2,57
3 Øygarden: Fana: Drammen: Stavanger: Molde: 2,76 Molde: 2,82 2,49 3,50 3,39 3,22 4 Dramme Eiganes: Eiganes: Midsund: Sunnmøre: n: 3,50 3,72 3,69 3,18 3,56 5 Arna: Sandnes: Sandnes: Fræna: 4,29 4,47 3,88 4,25 Sotra: Hå: 4,94 Hå: 5,28 11 12 4,55
2 12.03.2014
Order of voices .(Red=reg.centre, blue=national capital, green=rural region) Number of respondents
- Åsane - resp. who had moved to the locality after 6 yrs. of age - resp. who guessed that it was about dialect/language - resp. who did not tick off in all 8 x 15 = 120 evaluation scales
Midsund Molde Molde Ogna Stav StavSE Øygard Bergen Bergen Sum SE SE
52 72 72 33 71 61 80 176 165 782
13 14
Subconscious attitudes (pupils 15 yrs.) Mean ranks (A-tests = 2 x rural varieties)
Midsund Molde (N = Øygarden Bergen (N Hå Stavanger 4,5 (N = 52) 72) (N = 89) = 176 ) (N = 33) (N = 71) 4 1 Roms: 3,39 SEast: 3,77 SEast: 3,70 SEast : 3,86 Jær-y: 3,48 Jær-y: 3,82 3,5
2 Bg-high: Bg-high: 3 SEast: 3,24 Mo-low: 3,49 Jær-o: 3,21 Jær-o: 2,96 RegCLow 3,53 3,82 2,5 RegCHigh 3 Mo-low: Roms Sunn: Bg-low: Bg-low: Stav-high: Stav-high: RegRurOld 2 RegRurYoung 3,14 :2,70 2,70 2,88 3,17 2,88 2,95 SE/NCentre 4 Mo-high: Stril-y: Stav-low: 1,5 Sunn: 2,64 Stril-y: 2,21 SEast: 2,91 2,33 2,77 2,74 1 5 Mo-high: Stril-o: Stav-low: Stril-o: 1,94 SEast: 2,68 0,5 2,58 2,13 2,37 0 Midsund Molde Reg Øygarden Bergen Reg Hå Stavanger Reg
15 16
Ranking order of voices Mean rank (B-tests = 2 x SE varieties)
4,5
4
3,5
3
RwgCLow 2,5 RegCHigh RegRur 2 SE/NCentreLow SE/NCentreHigh 1,5
1
0,5
0 Molde Au Bergen Au Stavanger Au
17 18
3 12.03.2014
Ranking order of voices (lines connect identical voices) Significant contribution?
• Does the independent variable dialect make a significant contribution to the prediction of the variation in the dependent variable (is F Change sig <.05)?
19 20
The «noise» factors
Significant contribution? • Do other independent variables make a significant contribution to the prediction of the variation in the Test locality Dialect dependent variable (is F Change sig <.05)?
Molde SE NO Conclusion: Molde NO Dialect alone does not • Multiple linear hierarchical regression analysis Misund NO account for the Norwegian Stavanger SE YES
Stavanger NO variation
Hå NO
Bergen SE NO
Bergen NO
Øygarden NO
21 22
Reference to Reference to Speech geographic Playback Speech geographic Playback Test locality Age Intro Pitch Dialect Test locality Age Intro Pitch Dialect speed locality order speed locality order
Molde SE NO NO NO NO NO NO NO Molde SE NO NO NO NO NO NO NO
Molde NO NO NO NO NO NO Molde NO NO NO NO NO NO
Misund NO NO NO NO NO NO Misund NO NO NO NO NO NO
Stavanger SE YES YES NO NO NO NO YES Stavanger SE YES YES NO NO NO NO YES
Stavanger YES NO NO NO NO NO NO Stavanger YES NO NO NO NO NO NO
Hå NO NO NO NO NO NO NO Hå NO NO NO NO NO NO NO
Bergen SE YES YES YES NO NO NO NO Bergen SE YES YES YES NO NO NO NO
Bergen NO YES NO NO NO NO Bergen NO YES NO NO NO NO YES YES YES YES Øygarden NO NO NO NO NO Øygarden NO NO NO NO NO
23 24
4 12.03.2014
Pitch in the voices – and significant contribution (=S): Øygarden and Molde Reference to Speech geographic Playback Test locality Age Intro Pitch Dialect speed locality order
Molde SE NO NO NO NO NO NO NO S S Molde NO NO NO NO NO NO Misund NO NO NO NO NO NO S Stavanger SE YES YES NO NO NO NO YES S Stavanger YES NO NO NO NO NO NO Hå NO NO NO NO NO NO NO S Bergen SE YES YES YES NO NO NO NO
Bergen NO YES NO NO NO NO YES YES S Øygarden NO NO NO NO NO
Spearman’s rho .729**
25 26
Significant contributions from factors Reference to geographic Test locality Age Intro Pitch Speech speed locality Playback order Dialect Molde austlandsk .202 .112 .078 .001 .007 .023 .374 Molde sunnmørsk .161 .075 .006 .083 .10 .062 Age Intro Pitch Misund .019 .012 .058 .037 .04 .096 Stavanger austlandsk .41 .17 .014 .037 .017 .011 .328 Stavanger Jærbu .271 .038 .00 .006 .249 .002 .402 Stavanger Hå .024 .05 .004 .022 .172 .00 .703 (SE) Bergen austlandsk .246 .363 .198 .027 .015 .002 .029 Bergen stril .01 .633 .00 .00 .002 .21 Bergen (SE) Øygarden .01 .688 .00 .00 .022 .21
Bergen
Øygarden
27 28
Contribution from the various factors Conclusion
1 • «Noise factors» do better then dialect in accounting for
0,9 the evaluation (even though we tried to avoid extreme values of these factors) 0,8
0,7 Age 0,6 Intro Questions: Pitch 0,5 Speech speed • Is there a Norwegian pattern of ascribing values to Ref. to locality 0,4 voices? Playback order Dialect 0,3 • Are there regional patterns of ascribing values to voices?
0,2
0,1
0 Midsund Molde Molde SE Øygarden Bergen Bergen SE Hå Stavanger Stavanger SE
29 30
5 12.03.2014
Different national patterns w.r.t. determinant No national pattern in Norway. Perhaps a features/factors? regional one in Hordaland?
• 1
0,9
• 0,8 • 0,7 Age • 0,6 Intro Pitch • 0,5 Speech speed Ref. to locality 0,4 • Playback order Dialect 0,3 • 0,2 • 0,1
• 0 Midsund Molde Molde SE Øygarden Bergen Bergen SE Hå Stavanger Stavanger SE
31 32
Is Norway a ‘regionalized’ country? Do broadcast media play a decisive role? a) If TV plays a role, we should expect the same pattern Gallup poll (conscious attitudes) december 2013 (Sentio & all over the country (So far we’ve found different Nationen): patterns)
• N Norway: 61% like theirown dialect very much b) If radio plays a role, this may be different. (Many • Middle-Norway: 59% regional channels) • Oslo/Akershus: 38%
• N Norway: 66% veryproudoftheirowncounty • Middle-Norway: 70% • W Norway: 67% • Oslo/Akershus: 45%
33 34
Do broadcast media play a decisive role?
• Therefore: Of great interest to carry out VGTs in the • New question: other regions. From patterns of conscious attitudes we would expect SE Norway to be different. SE Norway has Are Norwegian subconscious attitudes formed by the great probably more bidialectalism and stylistic variation (= extent of exposure to linguistic variation? different language practice, less linguistic variation). (Use attitudes?)
• If also SE Norway turns out to be different w.r.t. subconscious attitudes, a) is wrong. • If SE Norway shows up a more «Danish» homogeneous pattern, this assumption has some support
35 36
6