Review of literature: Intonation in L2 teaching/learning
Bibliographies
Importance of intonation in L2: Theoretical studies
- Cantonese Speakers' Memory for English Sentences with Prosodic Cues
Martha C. Pennington & Nick C. Ellis, The Modern Language Journal, Volume 84, Issue 3, Page 372, Fall 2000.
- In the experiments, 30 Cantonese speakers with advanced competence in English were tested for their recognition memory of English sentences
in which prosody cued meaning contrasts in otherwise identical sentence pairs. The Cantonese speakers' memory for the English sentences
based on prosodic information was generally poor, both when the contrastive focus was implicit in the experimental task (Experiment 1)
and when it was the explicit focus of attention (Experiment 2). The only significant improvement in performance after participants' attention
was explicitly directed to intonation was on sentences in which prosody cued 'contrastive' vs. ' neutral' stress.
- Comment
The use of prosody in cuing various pragmatic meanings is difficult to recognize even for Cantonese speakers proficient in English.
Explicit attention to prosody improved the memory scores to a relatively small degree. Authors hypothesised that native speakers of tone languages
tend to store the intonational patterns holistically together with lexical material rather than analyze its component parts or meaning separately (see also Archibald 1997).
-
The structure and function of intonational paragraphs in native and nonnative speaker instructional discourse
Lucy Pickering, English for Specific Purposes, Volume 23, Issue 1, Pages 19-43 (2004).
- This paper examines the use of intonational paragraphs as an organizational tool in the teaching discourse of native speaker teaching assistants
and nonnative speaker international teaching assistants at a North American university. Six male North American TAs and six male Chinese (Mandarin) TAs were analyzed.
Analysis of the native speaker data presents evidence of intonational paragraphs defined by
phonological criteria and used by the speakers to underscore local and global information structure. Comparative analysis of the parallel nonnative speaker data shows decreased pitch range, a considerably
weaker control of intonational structure, disturbance in prosodic composition, all of which affects the comprehensibility of the discourse for native speaker hearers.
- Comment
Paper has a short description of Brazil's model of intonation in discourse (1997) that comprises three interacting systems - tone, key, and termination. "Key" and "termination"
divide the speaker’s pitch range into three levels: high, mid, and low. "Key" is especially useful since it marks if a given intonationall phrase is higher or lower relative to the preceding one
(marked with up and down arrows in Brazil's notation). The analysis in the paper, however, is only descriptive; no statistical analysis is provided.
-
Processing the prosody of oral presentations
Rebecca Hincks, Proceedings of InSTIL/ICALL2004 – NLP and Speech Technologies in Advanced Language Learning Systems – Venice 17-19 June, 2004.
- In this paper, speakers' ‘liveliness’ was measured as the Pitch Dynamic Quotient (PDQ, stdev divided by mean of F0). 20 native Swedish speakers delivered 10min
monologues in English. A high PDQ score correlated with the perception of enthusiastic and lively presentation while a low score correlated with boring, "deadpan" perception.
Mennen (1998) showed that te lack of confidence can manifest itself in a narrowed pitch range. However, dysfluencies also increse the PDQ score.
Therefore, some additional processing is needed.
- Acquisition of intonational prominence in English by Seoul Korean and Mandarin Chinese speakers
McGory, J. T., Unpublished Ph.D., Ohio State University, 1997.
- Reports a finding that the non-native speakers (Seoul Korean and Mandarin Chinese) did not distinguish between statements and questions in their f0 patterns.
The differences between non-native and native speakers of English could (to a large extent) be attributed to influences of the L1, and an effect of L1 background was found
in the different error patterns for the Mandarin and Korean speakers.
- The contribution of intonation to the perception of foreign accent
Jilka, M., Unpublished Ph.D., University of Stuttgart, 2000., Chapter 7 (conclussions)
- A comparison of intonational patterns in German and English of native American and German native speakers respectively. Many observations of Willems (1983) (Dutch vs. British English) are confirmed.
- Four major causes of intonational foreign accent were identified:
- a general incorrect choice and/or placement of tonal categories
- the transfer of tonal categories from the speaker’s L1 in corresponding discourse situations
- transfer in the phonetic realization of tonal categories
- overall characteristics such as relatively more tonal movement in the productions of the American speakers.
-
A study of sentence stress production in Mandarin speakers in American English
CHen, Y.- Robb, M.P.- Gilbert, H.R.- Lerman, J.W., Journal of the Acoustical Society of America 109, 4: 1681-1690, 2001.
- Bi-directional interference in the intonation of Dutch speakers of Greek
Mennen, I., Journal of Phonetics 32, 4: 543-563, 2004.
- The author found bi-directional interference in the realization of the prenuclear rising accent by Dutch speakers of Greek: not only does the L1 influence the L2, but the L2 also has an effect on the L1.
It was found that the majority of the L2 learners (four out of five learners) failed to produce native values of peak alignment in the L2, with a peak as early (i.e., within the accented vowel) as that of the native
Dutch control group, and considerably earlier than that of the native Greek control group (who realized the peak in the following unaccented vowel). Despite this lack of success in the phonetic realization
of the prenuclear rising accent in Greek, their realization of this rise in comparable Dutch utterances was nevertheless affected. Specifically, the four speakers were found to (partially) neutralize
the peak timing differences between the Dutch statements with long and those with short vowels in the accented syllable of the test word.
- Comment
this is a more 'phonetic' effect, nevertheless, it points to the difficulties caused by L1 <--> L2 transfer in the domain of intonation.
Teaching intonation with speech visualization technology
- no clear evidence that visual and auditory perception of F0 contours are directly related
- Jilka, M., see reference above
- An interesting application of speech technology: the use of rule-based F0 generation to produce alternative intonation contours to facilitate a subsequent
comparison of two versions of the same utterance by means of PSOLA-based resynthesis. The rule-based generation combines a system of intonation description and the generation mechanism proper,
which enables the user to deliberately alter the intonation contour and, with the help of the resynthesis, immediately evaluate these changes auditorily.
This procedure also has potential for future use in foreign language teaching. Learners could compare their own voices producing an inappropriate (original) intonation pattern
and a rule-generated alternative that is acceptable in the second language. The generation is controlled by easily interpretable ToBI tone labels and could therefore serve as the basis of pedagogical applications
of this kind. Another use of speech technology in this study is the application of synthesis (based on diphones from the MBROLA project) to generate stimuli without audible changes to the intonation contour
and/or without segmental foreign accent.
- Intonation Meter (A Visual Display for the Teaching of Intonation)
Gerard W. G. Spaai & Dik J. Hermes. Calico 10, 3 : 19-30, 1993.
- The authors discuss the use of an application that creates a stylized pitch contour and marks syllable onsets
to facilitate L2 intonation learning. The system was originally developed for deaf persons.
- Comment
Although the marking of syllable onsets might be helpful for learners (alignemnt of intonational events with paticular
syllables is important), the lack of transcription of the material in the display might be hindering the acquisition of this
alignment cues.
- Teaching Tone and Intonation with Microcimputers
D. Chun. Calico 7, 1 : 21-46, 1989.
- The article presents a review of research supporting the effectiveness of pitch visualization to facilitate
L2 intonation learning in 70s and 80s. It also reviews available (although now outdated) systems for doing so.
- Teaching intonation on questions
Susan Thompson, ELT Journal 49(3):235-243, 1995.