Selected Results
Data Collection, Method and Population
The FauvoT approach is a modified version of the regular audio-vocalic method established by Tomatis. Electronically modified audio-materials are used both for phases of concentrated listening that are integrated into regular pronunciation training in a language laboratory and for intensive listening in a home setting (for this, students are given a listening schedule and encouraged to listen regularly to the materials we uploaded on an e-learning platform).
Since 2007, the method has been used with university students at the University of Graz (advanced learners of English and beginners of Russian, Spanish, Italian and French), and with students of German at the University of Connecticut. The participants were divided into two groups (a FauvoT test group and a control group). The learners’ performance was recorded at the beginning and at the end of training (reading a short text or repeating words and phrases). In order to measure the learners’ progress, acoustic analyses of the vowels were carried out using the electronic software PRAAT (Boersma, 2001). Additionally, the students were asked to keep diaries and fill in questionnaires on their experience with the training method.
The FauvoT approach is a modified version of the regular audio-vocalic method established by Tomatis. Electronically modified audio-materials are used both for phases of concentrated listening that are integrated into regular pronunciation training in a language laboratory and for intensive listening in a home setting (for this, students are given a listening schedule and encouraged to listen regularly to the materials we uploaded on an e-learning platform).
Since 2007, the method has been used with university students at the University of Graz (advanced learners of English and beginners of Russian, Spanish, Italian and French), and with students of German at the University of Connecticut. The participants were divided into two groups (a FauvoT test group and a control group). The learners’ performance was recorded at the beginning and at the end of training (reading a short text or repeating words and phrases). In order to measure the learners’ progress, acoustic analyses of the vowels were carried out using the electronic software PRAAT (Boersma, 2001). Additionally, the students were asked to keep diaries and fill in questionnaires on their experience with the training method.
Evaluation of Data
The method is evaluated on several levels:
The method is evaluated on several levels:
- By means of an acoustic analysis of the recordings (before and after completion of the pronunciation training) using the acoustic speech software PRAAT
- By means of student questionnaires on the efficiency of the method
- By means of teacher questionnaires evaluating the efficiency and usefulness of the method
- By evaluating the listening diaries of the students
- By means of native speaker ratings evaluating the progress of the learners
Selected Results
The aim of our study was to find out whether the learners would be better able to approach the formant pattern of the native speakers after completing the training sessions. The hypothesis suggested that the group exposed to electronically modified audio material (group F) would come closer to the ideal formant structure than the control group that worked with non-modified material (group N).
Although not all of the parameters employed by Tomatis were used (no bone-conduction), the FauvoT groups achieved significantly better results than the control groups in our acoustic analysis of their pronunciation of vowels (see Eberl/ Pfandl 2008, Pfandl-Buchegger/ Eberl/ Landsiedler 2009, Pfandl/ Insam/ Landsiedler 2011a and 2011b).
Results for English
Results for British English (2010/11) show that the Austrian learners of English approached the vowel pattern of British English more closely after completing the training sessions, both in the overall form of the triangle and in the individual frequency values. As illustrated in the diagrams below, the form of the vowel triangle stretched towards the bottom in the FauvoT group and approached the more extended form of the British English native speakers (Figure 1a), while the vowel triangle of the controls shifted towards a different, more arbitrary form, unrelated to that of the native speakers (Figure 1b). With regard to individual sounds, the much better approximation of the vowel /æ/, which remains in its restricted position in the control group, as is more characteristic of (Austrian) German, is especially noticeable (see Figures 1a and 1b).
Although not all of the parameters employed by Tomatis were used (no bone-conduction), the FauvoT groups achieved significantly better results than the control groups in our acoustic analysis of their pronunciation of vowels (see Eberl/ Pfandl 2008, Pfandl-Buchegger/ Eberl/ Landsiedler 2009, Pfandl/ Insam/ Landsiedler 2011a and 2011b).
Results for English
Results for British English (2010/11) show that the Austrian learners of English approached the vowel pattern of British English more closely after completing the training sessions, both in the overall form of the triangle and in the individual frequency values. As illustrated in the diagrams below, the form of the vowel triangle stretched towards the bottom in the FauvoT group and approached the more extended form of the British English native speakers (Figure 1a), while the vowel triangle of the controls shifted towards a different, more arbitrary form, unrelated to that of the native speakers (Figure 1b). With regard to individual sounds, the much better approximation of the vowel /æ/, which remains in its restricted position in the control group, as is more characteristic of (Austrian) German, is especially noticeable (see Figures 1a and 1b).
Figure 1a: Formants of the vowels /i/, /æ/, and /u/ in the FauvoT group: first reading (brown triangle), second reading (maroon triangle), British native speakers (blue triangle)
Figure 1b: Results of the control group
Results from Russian
Similar results were achieved by Austrian students of Russian (beginners) after completing an introductory pronunciation class. The experimental group was able to reproduce the Russian formant pattern more closely than the controls which suggests that they were more successful in imitating the characteristic features of the Russian articulation patterns in the sounds tested..
Figure 2a illustrates the vowel positions (in correlation to their articulatory setting) in the initial student recordings before their first training session reveals significant differences in the vowel space used by Russian native speakers and the Austrian students of Russian in terms of the mean frequency patterns of the first two formants, F1 and F2. It can be inferred from the diagram that Russian native speakers generally produce vowels at more extreme ends of the vowel diagram, while the students primarily used the more central vowel space that is determined by their native language, Austrian German.
Figure 2b contrasts the vowel space of Russian native speakers (outer dotted square), the FauvoT-group (inner dark square), and the control group (inner light square) on completion of the pronunciation class, after two weeks of concentrated listening sessions.
Similar results were achieved by Austrian students of Russian (beginners) after completing an introductory pronunciation class. The experimental group was able to reproduce the Russian formant pattern more closely than the controls which suggests that they were more successful in imitating the characteristic features of the Russian articulation patterns in the sounds tested..
Figure 2a illustrates the vowel positions (in correlation to their articulatory setting) in the initial student recordings before their first training session reveals significant differences in the vowel space used by Russian native speakers and the Austrian students of Russian in terms of the mean frequency patterns of the first two formants, F1 and F2. It can be inferred from the diagram that Russian native speakers generally produce vowels at more extreme ends of the vowel diagram, while the students primarily used the more central vowel space that is determined by their native language, Austrian German.
Figure 2b contrasts the vowel space of Russian native speakers (outer dotted square), the FauvoT-group (inner dark square), and the control group (inner light square) on completion of the pronunciation class, after two weeks of concentrated listening sessions.
Figure 2a: Russian vowel space (blue), Austrian learners of Russian (red dotted line).
1st recording. Vowels in the words (left to right): Mila, bufet, sof’ja, papa, etaž.
1st recording. Vowels in the words (left to right): Mila, bufet, sof’ja, papa, etaž.
Although both groups were able to approach the Russian vowel pattern, the FauvoT group approached the overall structure of the Russian vowel pattern more closely than the controls and managed better to reproduce the positions of the higher front vowels /i/ and /e/ and to approximate the Russian formant frequencies of the back vowel /u/. The test group largely retained the formant frequencies of the vowel that are more typical of Austrian German.
Figure 2b: Vowel space of Russian vowels, 2nd recording: FauvoT-group (red), control-group (green), Russian native speakers (blue). Vowels in the words (left to right): Mila, bufet, papa, etaž.
For a more detailled analysis see the report on Russian:
For a more detailled analysis see the report on Russian:
bericht_russisch.pdf | |
File Size: | 2413 kb |
File Type: |
Improved student perception:
For improved student perception in the wake of intensive pronunciation training see the results of a sound discrimination test (Pfandl-Buchegger and Insam forthcoming) with advanced learners of English showing a high percentage (72%) of improvement in the perception of English speech sounds before and after pronunciation training.
Native speaker ratings :
Native speaker ratings (Eberl/Pfandl-Buchegger 2008) were used to examine the effect of the so-called ‘audio-vocalic loop’ (i.e. electronically modified feedback of the speaker’s voice). They found a significant improvement (68%) in the overall impression, most noticeably in the realm of supra-segmental characteristics such as intonation and rhythm (65%) and commented on the greater fluency, accuracy and intelligibility of the readings with modified feedback.
For improved student perception in the wake of intensive pronunciation training see the results of a sound discrimination test (Pfandl-Buchegger and Insam forthcoming) with advanced learners of English showing a high percentage (72%) of improvement in the perception of English speech sounds before and after pronunciation training.
Native speaker ratings :
Native speaker ratings (Eberl/Pfandl-Buchegger 2008) were used to examine the effect of the so-called ‘audio-vocalic loop’ (i.e. electronically modified feedback of the speaker’s voice). They found a significant improvement (68%) in the overall impression, most noticeably in the realm of supra-segmental characteristics such as intonation and rhythm (65%) and commented on the greater fluency, accuracy and intelligibility of the readings with modified feedback.
Improvement: yes | 77.35% | |
---|---|---|
Overall impression: | 68% improved | |
Improved | ||
noticeably | well noticeably | |
53% | 16% | |
Intonation | 56% | 14% |
Stress and rhythm | 56% | 13% |
Consonants | 44% | 6% |
Vowels | 34% | 8% |
Similar improvements were obtained by a study testing the feedback loop with students of Russian (see Auckenthaler, diploma thesis 2010). Additionally, native speaker ratings on intonation in the project FauvoT (Insam, in progress) attested significantly higher scores for the experimental group after the training period (61%) than for the control group (42%).
Student Questionnaires:
Results from the student evaluation within the project FavoT (diaries and questionnaires) also show favourable reactions (mainly in terms of motivation): the attested significant increase (83%) in confidence obviously results in higher motivation to speak and an increased use of the L2 (Pfandl 2008, Landsiedler/ Pfandl/ Insam 2009).
According to the evaluated material, students are more willing to use the target language after the training sessions. All in all, their motivation and pleasure in learning the new language has increased.
Results from the student evaluation within the project FavoT (diaries and questionnaires) also show favourable reactions (mainly in terms of motivation): the attested significant increase (83%) in confidence obviously results in higher motivation to speak and an increased use of the L2 (Pfandl 2008, Landsiedler/ Pfandl/ Insam 2009).
According to the evaluated material, students are more willing to use the target language after the training sessions. All in all, their motivation and pleasure in learning the new language has increased.
Confidence when using foreign language | |
---|---|
Higher | 83,3% |
Less | 8,3% |
No change | 8,3% |
Increase in use of foreign language | |
Yes, more | 25% |
A little more | 50% |
No | 25% |
Degree of motivation and pleasure | |
Higher | 75% |
Less | 0% |
No influence | 25% |
For a more detailed analysis of the questionnaires see Landsiedler/Pfandl-Buchegger/Insam (2011).
Teacher Feedback:
Feedback from a teacher-training course (Pfandl 2008) confirmed a more accelerated perception of sounds that also led to a faster memorization of foreign language words. Additionally, teacher feedback on classroom experience with pronunciation teaching tends to confirm our obser¬vations.
For more detailed results see
Link: Projektbericht 2008
Feedback from a teacher-training course (Pfandl 2008) confirmed a more accelerated perception of sounds that also led to a faster memorization of foreign language words. Additionally, teacher feedback on classroom experience with pronunciation teaching tends to confirm our obser¬vations.
For more detailed results see
Link: Projektbericht 2008