Bugge: attitudes to variation in spoken Faroese


Method (overview)

  • Adapted from method used in Western Norway study
  • 15 voice samples with 15 seconds of speech on the same topic
  • 3 15-year-old females from 5 different regions
  • Topic: G! Festival
  • Test taken in 7 different schools in Faroes
  • Stratified sampling: the researcher divides the population into regional groups, called strata. Then, a probability sample (often a simple random sample) is drawn from each regional group.
  • 203 students (1/3 of 15 year olds at the time) (could have tested adults however they are more suspicious of the test purpose as not subject to random tests more frequently/recently as students are)

1. VERBAL GUISE TEST: students asked to rate voice samples against 8 personality traits on a Likert scale

2. Students asked to guess test purpose

3. Student asked to state where they think guises are from (if they failed to guess correctly, their data was not included in the analysis)

4. LABEL RANKING TASK: students asked to rank dialects according to 1. Their beauty 2. Their status

5. Data on test takers social background collected

1 of 17

Verbal Guise Test Advantages

  • Reveals covert attitudes
  • Guise anonymity ensures personal variables are controlled so that guises remain strangers/impersonal representatives of the language variety
  • Objective attitudes to the dialects as opposed to the person
  • Topic chosen is as neutral as possible
  • Guises all female so gender variable is controlle
  • Largely representative sample of test takers (1/3 of all 15 year olds in the Faroes at the time)
2 of 17

Verbal Guise Test Disadvantages

  • Experimental test: artificiality, not naturalistic
  • Monostylistic presupposition of dialectal variety
  • Guises may not sufficiently represent the dialect
  • Guises may encourage stereotypes
  • Guises all said different things so test takers attitudes might be based on this as opposed to dialect
  • Topic not completely neutral: religious students may deem the festival immoral, other students may see the test as a device for teachers/parents to find out if they'll be drinking alcohol
3 of 17

Validity of Verbal Guise Test

Concealment of Guise Test Purpose:

  • No students or teachers were told/guessed the test purpose
  • Test administered simultaneously - reduces risk of test purpose spreading
  • Fieldworkers sent to regions they didn't grow up in
  • Study on students - who are subject to random tests more frequently and recently than adults who might have been more suspicious of the test purpose

Maintenance of Guise Anonymity

  • Guises need to be anonymous representations of the dialects
  • Personal attitudes need to be controlled to maintain anonymity and objective attitudes
  • 3 year time gap between collection of speech samples and administration of the test
  • No students guessed correctly the identity of the guise
  • However, some students noted a personal name: perhaps reduces the validity as the test turns in to a 'Guessing Game' as opposed to eliciting attitudes to dialect
  • Due to small tight knit communities: everyone knows everyone
4 of 17

Alternative Method to Verbal Guise Test

  • Matched Guise Test
  • Test takers listen to apparently different speakers representing language. They evaluate the speakers according to various traits. The speaker is actually bilingual or polyglot.
  • However, test takers may be aware its only one bilingual/polyglot individual and not different representatives of the varieties
5 of 17

Likert Scale Advantages

  • Gives options
  • Doesn't force the test takers to take a particular stand
  • Accomodates neutral and undecided feelings
  • Qualitative attitude responses are easily quantifiable
  • Versatile and inexpensice
6 of 17

Likert Scale Disadvantages

  • Uni-dimensional: only gives 5-7 options (Bugge doesn't state how many he uses)
  • People may avoid choosing the 'extremes' because of negative implications involved with 'extremists'
  • Attitudes may not be truthful
  • Traits may not sufficiently reflect the language/socio-economic discourse/cultural stereotypes associated with a dialect
  • Likewise the traits may perpetuate stereotypes
7 of 17

Alternative Method to Likert Scale

  • Thurstone scale
  • Attitude measurement
  • Made up of statements about a particular issue, and each statement has a numerical value indicating how favorable or unfavorable it is judged to be. People check each of the statements to which they agree, and a mean score is computed, indicating their attitude.
  • However, again test takers may avoid choosing extremes
  • Statements may also encourage stereotypes
8 of 17

Verbal Guise Test Results

  • Reaction patterns reflect no dialectal status hierarchy
  • inferential tests performed: ANOVA
  • P > 0.05 so data not significiant (dialect variable not influencing students evaluation pattern)
  • However, significant correlation between traits
  • Evaluation patterns distributed along dimensions of SUPERIORITY and DYNAMISM
  • The variable leading to this correlation is unknown
  • The test takers that failed to guess where the guise was from were not included in the analysis
9 of 17

Label Ranking Task Advantages

  • Elicits overt attitudes to dialect
  • Inexpensive & versatile
10 of 17

Label Ranking Task Disadvantages

  • Experimental test: artificiality & not naturalistic
  • Pre-defined hierarchy
  • Forced to put dialects in order - no option for neutrality
  • Elicits pre-concieved/socially acceptable ideas of dialect status - perhaps not the test takers truthful personal opinion
11 of 17

Label Ranking Test Results

  • Dialect given the highest beauty ranking is the dialect of the region the test takers school is situated in
  • Torshavn is ranked as the highest status
  • Sudoroy is ranked as the lowest status
  • Other regions ranked according to their relative distance away from Torshavn (the capital)
12 of 17

Verbal Guise Test Results Analysis

  • Evaluation pattern not random but not linked to attitudes towards dialects
  • Results revealed no status hierarchy
  • This could reflect a covert verbal tolerance of regional dialect
  • As well as private attitudes and a hidden prestige of certain varieties
  • This is a different pattern found in the results of the Danish study
  • Danish dialectal features are more socially loaded
  • Danish people are more likely to be triggered by dialect usage
  • Faroes and Western Norway use dialect more frequently
13 of 17

Label Ranking Task Results Analysis

  • Preference for own dialect also found in Danish results
  • Pattern of Torshavn as the highest status and Sudoroy as the lowest status mirrors the political and financial discourse known from Faroese political discourse
  • Torshavn is the power centre
  • Results reflect the standard ideology of assigned status to dialectal varieties
  • Same as the Danish pattern - upgrading of the capital dialect
14 of 17

Comparison of Verbal Guise Test and Label Ranking

  • Verbal guise test didn't reveal a status hierarchy
  • The label ranking test did
  • Verbal guise test results reveal covert attitudes, tolerance and a hidden prestige
  • Label ranking results reflect the standard ideology and accepted assigned status to dialectal varieties
  • ALTERNATIVELY: differences in evaluation could be down to the mixed methods approach and the two different types of stimuli
  • Verbal Guise Test uses speech sample
  • Likert Scale requires responses to a list of pre-defined traits
  • Words may not sufficiently reflect socio-economic/language discourse/cultural sterotypes
15 of 17

Bugge's Epistimology


  • Society shapes the individual
  • Social facts exercise coercive control over individuals
  • Social norms govern action
  • Label Ranking Results (which revealed a status hierarchy): language ideology (a social construct) as an explanation
  • School/Education as a force: Eivind Weyhe (2015) states current reading style in schools favours the phonology of Central Faroese dialects
  • Media as a force: radio broadcasts favoured Faroese dialects from the populous southern part of the central island Streymoy
  • However all regional varieties present in broadcast
16 of 17

Bugge's Epistomology


  • Some evidence of an interpretivist approach
  • Recognises differences between test takers: not just social products
  • Likert scale: labels may not be related to or reflect the test takers language discourse/socio-economic/cultural sterotypes
  • Verbal Guise Test: the test takers may not regard the guises as valid representatives of the cultural or socio-economic stereotypes or the socially constructed dialect norms
  • There is no OBJECTIVE REALITY as positivism suggests
17 of 17


No comments have yet been made

Similar English resources:

See all English resources »See all Research Methods resources »