<<

How to choose between pronunciation goals when teaching Swedish as a second Bosse Thorén Dept. of Language Studies, Umeå University

High and realistic goals, suitable for adult learners with different capacities. How far will a strict linguistic-phonological analysis take us in communicative and educational reality? If a native-like L2-pronunciation is the only with different L1’s. The suggestions for core The overview below shows 6 acceptable learning target, every phonological and features in Swedish are based on columns. Each phonological phonetic detail is of equal importance. Adult L2- empirical studies, structural circumstances and contrast can however be learners seldom reach this goal. This has been teaching experience. The suggestions include associated with a combination of discussed in the case of English by e.g. Jenkins considerations as to what phonetic/acoustic details acoustic cues. On the phonetic (2000, 2002), for Swedish by Bannert (1984). play a role in realizing the phonological contrasts. level there is often a combination of Jenkins introduces the Lingua Franca Core, i.e. In older didactic tradition, the features that were spectral, temporal and tonal phonological and phonetic features assumed to be hard to learn got the highest priority and invest- features, signaling the contrast. more crucial to intelligibility than others. Swedish is ment in time, regardless of their importance to L2-speakers may perceive and not a lingua franca in a global sense, but is used in communication (provided that pronunciation was produce a phonological contrast, Sweden and in Finland as a means of commu- taught at all). Among such features are the tonal but using other phonetic cues or nication between native speakers with different word accents and a voiceless [ɧ], that can combination of cues, resulting in a regional accents and between native and non- be substituted by [ʃ] or [ʂ] with maintained clarity. foreign accent native speakers and among non-native speakers And , vowels, vowels…

Overview of Swedish Phonology Segments / phonemes Phonotactics Length/ / Vowels Consonants and phonological quantity intonation processes

Distinctive stress on A quantity distinction A tonal word accent 9 phonemes Voiced 0-3 consonants initially. If 3, the word level: involving comple- contrast, realized in /a u ʉ o e i y ɛ /b/ /d/ /g/ /j/ /l/ /m/ first must be /s/, the second a ˈJapan – jaˈpan mentary vowel and higher prominence ø/. / y ʉ ø/ are /n/ /r/ /v/ /ŋ/ voiceless stop, and the third a consonant duration categories voiced continuant. ‘Japan’ – ‘Japanese’ front rounded. Voiceless as well as spectral (accentuated and ˈRacket – raˈket Quality (timbre) /f/ /h/ /k/ /p/ /s/ /t/ 0-5 consonants in final position, difference between sentence/focal difference between /ɧ/ /ɕ/ but examples are few. 3-4 is ‘racket’ – ‘rocket’ long and short vowel accent) long and short common. allophone Accent 1 (acute): Stops vowel allophone; A de-voicing changes On phrase level: tal – tall /b/ /d/ /ɡ/ /k/ /p/ /t/ ánden ’the duck’ Substantial for many voiced obstruents into ‘number’ – ‘pine’ ˈHälsa på någon Accent 2 (grave) /a/, /ʉ/ in most /f/ /h/ /j/ voiceless allophones. ’to greet someone’ ful – full varieties, negligi- /s/ /ɧ/* /ɕ/ /v/ /r/ + dental  supradentals ‘ugly’ – ‘full/drunk’ ànden ’the spirit’ ble for /ɛ ø/, and Hälsa ˈpå någon Nasals /m/ /n/ /ŋ/ /n/ adapts place of articulation to virtually absent in ’to visit someone’ etc. following obstruent: [ŋ] [ɱ] [m] [ɲ] Finland-Swedish. Liquids /l/ /r/ Some standardized reductions.

Stress is a clear candidate to core Tonal word accent is formally Phonemes; vowels and consonants. Realizing consonants and feature. It is the only phonemic contrast phonemic in those varieties The front rounded vowels /y ʉ ø/ are consonant clusters is crucial (in Swedish) that is empirically proven utilizing it, but proven to be of known to be rare in the worlds for intelligibility. Recent crucial for intelligibility (Bannert 1987 + a less importance to intelligibility . From experience we can tell immigration from Vietnam, host of anecdotal evidence). in practice. It has several that L2-speakers make themselves Thailand, China and Korea Stress interpreted as promi- regional kinds of realization understood without discerning all 9 has shown that distorted nence may be signaled by dynamic, and is neutralized in singing. vowel phonemes. The exact vowel consonant clusters often temporal, spectral and tonal means. The Tones in Swedish do proba- quality varies between regions, as does compromise intelligibility. most reliable perceptual cue to Swedish bly have a more important role the quality difference between long and Abrahamsson (2013) and syllable prominence is increased in telling the listener which short allophone. Jenkins (2002, for English) duration of stressed (Fant & words are “news” (rising pitch) The consonants are greater in number claim that consonant Kruckenberg 1994). This finding is a and when the utterance is and also in possible combinations. A text deletion is more harmful good reason to combine stress completed (falling pitch). where vowels are removed (or replaced than vowel epenthesis, as realization with the strongly duration- No distinct tonal patterns for by e.g. an asterisk) is much easier to means of reducing dependent quantity contrast in education statement or question. read than a text where consonants are consonant clusters. (Thorén 2008) removed.

Conclusion References General: give priority to features shared by most regional varieties and not to ‘local Abrahamsson, N. (2013). Fonologiska aspekter på andraspråksinlärning och svenska som andraspråk. Ingår i Hyltenstam, K. & Lindberg, I. (red) Svenska som specialties’ even though they might be part of a high status variety. andraspråk i forskning, undervisning och samhälle. Lund: Studentlitteratur Segments: Consonant realization and intact consonant clusters are more crucial to Bannert, R (1984), Prosody and intelligibility of Swedish spoken with a foreign accent. Nordic Prosody III. Acta Universitatis Umensis. Umeå University. intelligibility than vowel quality. Bannert, R. (1987) From prominent syllables to a skeleton of meaning: a model of Prosody: Stress and quantity are more crucial than the tonal word accents. Stress and prosodically guided speech recognition. In Proceedings of the XIth ICPhS Tallinn, quantity can be assumed to provide a temporal-rhythmical base for intelligible Swedish. pp 73-76. Fant G. Kruckenberg A. (1994) Notes on stress and word accent in Swedish STL- In didactic practice, the phonetic feature of duration is a very convenient means of QPSR 2-3/1994 promoting both the phonological contrasts of stress and quantity, since it is the main Jenkins, J. (2000). The phonology of English as an international acoustic correlate (perceptual cue) to both. The teacher can kill two birds with one stone. language. Oxford: Oxford University Press. Jenkins, J. (2002). A Sociolinguistically Based, Empirically Researched For an effective pronunciation teaching, one must take into account phonological factors, Pronunciation Syllabus for English as an International language. Applied Linguistics the relative weights of phonetic/acoustic cues in realizing the phonological contrasts, as 23/1: 83-103 well as knowledge about which features are easier and harder to learn and which features Thorén, B. (2008) The priority of temporal aspects in L2-Swedish prosody: Studies in perception and production. Doctoral thesis, dept of Linguistics, are crucial for intelligibility. More research in this area is required. University.