4.2 Investigated speakers and word choice

All the 16 participants investigated are native speakers of German (13 female, 3 male) with English as a second language. In addition, they are all students of English at the University of Bamberg. After the recording, each speaker had to state whether he/she aims at BrE or AmE as a model for his/her pronunciation. A majority of speakers (62.5 %) preferred American pronunciation. 37.5 % of the speakers preferred British pronunciation.

For the analysis, 6 words were chosen. All the chosen words have a different word stress in RP and GA. Only such words were included in which the syllables in question are short when unstressed, which excludes a distorting effect of phonemic vowel length. There are a lot of words that have varying word stress in RP and GA, but in most cases, there are for example two variants of word stress in GA, while one of these variants is the typical pronunciation in RP. Take for instance the word ballet which is pronounced /’bæleɪ/ in RP and /bæl’eɪ/ or /’bæleɪ/ in GA. If the speaker stresses the first syllable of this word, it cannot be determined if it is the RP or GA variant of the word. Thus, with such words it is impossible to decide if a speaker is consistent in his or her pronunciation. This conclusion can be drawn more clearly if there is just one possible stress placement in each GA and RP. This is the case for the following words:

Figure 4: Chosen Words for Mini-project with Varying Stress Pattern in RP and GA (acc. to Jones 2011)

The pronunciations were checked with the help of the Cambridge English Pronouncing Dictionary (18th ed.). One weakness of the word choice is that some words are not in common use and might prove to be difficult for the speakers, for example miscellany, ancillary or capillary. These words belong to an advanced vocabulary.

The speakers had to read these words out loud and they were recorded with the help of the software Praat. Afterwards, I used both the auditory and the acoustic method for analysis. With the auditory method, the stress placement of the words was detected by simply listening. In order to receive more objective and reliable results, I listened to the recordings several times. For a larger-scale project, consulting a second rater would  increase objectivity. For the acoustic method, I used the software Praat and indicated the stress placement by analysing the features vowel duration, intensity and pitch height.

Created with the Personal Edition of HelpNDoc: Easy EPub and documentation editor