The following post was contributed by Veronika Thir. You can read a short biography of Veronika and her work at the end of the post.
What happens when ELF users try to understand each other’s accents?
Language teachers are often painfully aware of the limited time they have in the classroom to equip their learners with what they need for international communication. Therefore, they are always interested in knowing what aspect of language (use) they should focus on most in their lessons. Grammar? Vocabulary? Pronunciation? Communication strategies?
In the past, some claims have been made that pronunciation seems to be a particularly serious ‘problem area’ for users of ELF in that it turned out to be the major source of communication problems in some studies (Jenkins 2000, Deterding 2013). Accordingly, calls have been made for a stronger focus on pronunciation in the ELT classroom. However, what has often been overlooked is that a couple of other studies (e.g. Mauranen (2006), Pitzl (2010) or Kaur (2011)) did not actually find pronunciation to be a serious problem for mutual understanding in ELF communication. It seems that pronunciation might sometimes be responsible for the vast majority of communication problems in ELF talk, while sometimes, it is not much of a problem at all. But how can we explain this variable role of pronunciation for successful ELF communication?
One possible explanation is the availability of co-text and context in certain situations. If we process each other’s pronunciation to recognize words in the stream of speech, we do not only draw on the sounds we hear (‘bottom-up’ processing). We also use information from the linguistic co-text and the extra-linguistic context to make sense of what we hear (‘top-down’ processing).
Let me provide an example: if we hear someone saying “I [s]ink this is a nice idea”, we will most probably not assume that the person intends to say ‘sink’ but that they mean ‘think’, because the lexical and syntactic co-text do not support our interpretation of ‘sink’. Possibly, the situational context will also provide us with certain cues that lead us to prefer ‘think’ over ‘sink’ in our interpretation. We might also consider our knowledge of the speaker (e.g. that they often replace ‘th’ with ‘s’) and use this information to interpret what they say, and conclude that ‘think’ is the intended word.
What I’m getting at is that both co-text and context can supply the listener with cues that might help them to compensate for ambiguities in the acoustic signal. This is why Brown (1989) called context “a powerful disambiguator“. If co-textual and/or contextual cues are sparse or absent, a certain accent might quickly become much more of a problem than it would have been otherwise.
The interesting thing is that we do not yet really know how much of an impact co-textual and contextual cues have on the intelligibility of pronunciation in ELF communication. Jennifer Jenkins (2000) observed that ELF users sometimes seem to be unable to profit from such cues when trying to understand one another. However, other ELF researchers have suggested that co-text and context might very well be an important additional source of information for ELF listeners in understanding their interlocutor’s accent (Deterding (2012), Luchini & Kennedy (2013), Osimk (2009)).
My doctoral research aims to shed more light on the role of co-textual and contextual cues when ELF users process each other’s accents. To this end, I recently carried out a pilot study, part of which I presented at ELF10.
What did I do?
I gathered data from 2 pairs of ELF speakers who completed two communicative tasks under different conditions: the experimental condition, which involved the presence of a certain ‘schematic’ context which participants were able to draw on in order to make sense of their interlocutor’s speech, and the control condition, in which no such context was available.
Each pair completed one of the two tasks (map-task and spot-the-difference task) under the experimental condition and one task under the control condition. All interactions were video-recorded and each participant re-watched the video of the interactions they had taken part in together with me in order to help me identify all instances of phonological intelligibility problems in the data and clarify their sources. I then analysed each instance qualitatively, in particular with regard to the presence or absence of schematic or visual context and linguistic co-text.
What did I find?
So here are the most important observations I was able to make regarding the 2 pairs of ELF speakers I examined:
- The ELF speakers in this study did use co-textual and contextual cues in understanding each other’s accents. They did so not only when hearing something for the first time, but also when trying to resolve intelligibility problems and negotiating meaning. Apart from linguistic co-text, visual context (e.g. in the form of the map in front of them, or their partner’s body language) and verbal cues evoking part of a relevant schema (e.g. “the place with animals” for ‘zoo’) turned out to be helpful in resolving comprehension difficulties.
- Most of the time they were used, co-textual and contextual cues turned out to be helpful to ELF users, in that they aided them in correctly identifying the word(s) their partner was uttering. However, in some cases such cues were unhelpful, in that they led the listener in the wrong direction by supporting an incorrect interpretation of their partner’s pronunciation.
- In the absence of co-textual and contextual cues, there seems to be a greater danger of misunderstanding. The few true misunderstandings in my data – i.e. instances where the listeners thought they had understood the word their partner was uttering when in fact they had not – all occurred in the control condition without schematic context. In each case, the reason the listener thought they had understood correctly seemed to be that there were no co-textual or contextual cues that would have led them to question their understanding – their interpretation of their partner’s pronunciation seemed accurate during the task. This suggests that a lack of co-text and context can be detrimental to communicative success in that it prevents ELF users from noticing a problem and, therefore, clarifying it.
Possible implications for teaching
It goes without saying that the results of my research are tentative, as they are based on a very small number of participants. Far more research, not only of a qualitative but also of a quantitative nature, is needed before the exact implications for English language teaching regarding the interplay of co-text, context and pronunciation in ELF communication can be identified. Rather than giving definite recommendations for teaching, I would therefore like to provide an outlook for what such research might find, along with its possible implications:
- Research might find that the presence of linguistic co-text and/or different forms of extra-linguistic context is indeed important for understanding another ELF user’s accent. If this is the case, learners might have to practice:
- producing sufficient co-text for their listeners to draw on;
- ‘triggering’ relevant schemata in their listeners (e.g. by using body language or by verbally creating associations)
- drawing their listeners’ attention to the physical context (again e.g. by using body language).
All of these strategies might help learners to resolve or prevent phonological intelligibility problems in ELF talk if they are unable to resolve or prevent them otherwise, e.g. by getting a particular sound right.
- Possibly, certain contexts of real-life language use might be identified that maximize the risk of phonological intelligibility problems in ELF talk in contrast to those that minimize them. In this case, it will be important to raise learners’ awareness of the contexts in which they need to either pay particular attention to their pronunciation or employ the communication strategies mentioned above to prevent intelligibility problems.
However, as mentioned above, a lot more research is still needed to further explore and confirm (or not) the tendencies I observed in my study. I’m on it! So, watch this space!
 This ‘interactive’ view on processing speech has been supported by research findings in psycholinguistics (see e.g. Byrd & Mintz 2010: 162) and on the intelligibility of L2 speech (Zielinski 2006).
 These 2 pairs were selected from a pool of 5 pairs participating in this study, as they turned out to be fairly comparable in terms of certain extra-linguistic factors that have been found to affect intelligibility, such as familiarity with their partner’s accent and language attitudes (this information was obtained through a follow-up questionnaire). However, they were not comparable in terms of production, in that the strength of their L1 accent was different. It is therefore problematic to compare the 2 pairs directly to each other, which is why I decided to focus on comparing the 2 task conditions rather than the 2 pairs.
Brown, Adam. 1989. “Some thoughts on intelligibility”. The English Teacher XVIII. http://www.melta.org.my/ET/1989/main4.html (January 16, 2016).
Byrd, Dani; Mintz, Toben H. 2010. Discovering speech, words, and mind. Malden, Mass: Wiley-Blackwell.
Deterding, David. 2012. “Intelligibility in spoken ELF”. Journal of English as a Lingua Franca 1(1), 185–190.
Deterding, David. 2013. Misunderstandings in English as a Lingua Franca. Berlin: de Gruyter Mouton.
Jenkins, Jennifer. 2000. The phonology of English as an international language. Oxford: Oxford University Press.
Kaur, Jagdish. 2011. “Intercultural communication in English as a lingua franca”. Intercultural Pragmatics 8(1), 93–116.
Luchini, Pedro L.; Kennedy, Sara. 2013. “Exploring sources of phonological unintelligibility in spontaneous speech”. International Journal of English and Literature 4(3), 79–88.
Mauranen, Anna. 2006. “Signaling and preventing misunderstanding in English as lingua franca communication”. International Journal of the Sociology of Language 2006(177), 123–150.
Osimk, Ruth. 2009. “Decoding sounds”. Vienna English Working Papers 18(1), 64–89.
Pitzl, Marie-Luise. 2010. English as a lingua franca in international business. Saarbrücken: VDM Müller.
Zielinski, Beth. 2006. Reduced intelligibility in L2 speakers of English. Bundoora, Victoria.
About the author: Veronika Thir
Veronika Thir is a university assistant and PhD student at the English Department of the University of Vienna, where she obtained a teaching degree in English and French. During her studies, she has worked as a student tutor for practical English phonetics. While her MA thesis focused on the pedagogical implications of ELF research for teaching English pronunciation at university level (see also her article in VIEWS, 2016), her PhD project seeks to contribute to research on phonological intelligibility in ELF by exploring the interplay of pronunciation, co-text and context. She has also presented at various academic conferences.