Hot Topics in Prague

HOT TOPICS IN PRAGUE

The 4th International Conference on English Pronunciation: Issues & Practices (EPIP) was a three day event in Prague. EPIP conferences occur at two-yearly intervals, and previous editions had taken place in France, South Africa and Spain. The conference was attended by just under 100 participants, many of whom worked in an academic environment such as the one which hosted the event – the Institute of Phonetics of Prague’s Charles University. The event was organised by Jan Volin, Kristyna Poesová, Lenka Weingartová and Radek Skarnitzl. (Jan and Lenka were also part of the swing quintet that provided our entertainment at the conference dinner!). Below are a selection of the topics discussed during the conference, and some of the presentations that I was able to attend.

Pataka

How are we to distinguish the consonants /p/, /t/ and /k/ from /b/, /d/ and /g/? Well, voicing, right? Not so simple. Actually, we need to consider if the sound is at the beginning or end of the syllable. If it’s at the beginning, the /p/, /t/ and /k/ are characterised by aspiration – the small explosion of air from the mouth which will put out a candle if needs be. This is lacking for /b/, /d/ and /g/. If the sound is at the end of the syllable, then the thing to focus on is the vowel sound before it. /p/, /t/ and /k/ (and other unvoiced consonants) will tend to cut the vowel a bit short, and this may be a greater auditory clue than the voicing itself. This issue came up in a number of presentations in Prague – so much so that people were using a new word as shorthand for it: pataka (the letters P, T and K joined up with schwas).  

Marta Zając took an unusual angle on the topic. She reported on an experiment to see to what extent Polish speakers of English would vary their pronunciation of these consonants according to whether they were speaking to another Pole or to a native speaker of English. The design of the experiment was fiendishly clever, accommodation being a very tricky thing to measure. The results seem to indicate that when speaking to an English native speaker, the subject would ‘beef up’ the initial aspiration in order to sound more English. They didn’t beef up their voicing, however – perhaps because this is more difficult or less salient for them.

Bojana Jakovljević and Maja Marković chose to focus on the vowel-shortening effect of the unvoiced final consonants (pre-fortis clipping), and the extent to which Serbian learners of English acquire this feature. They pointed out that this shortening effect is in part physiologically controlled – we do it because we can’t do otherwise. To this extent, Serbian learners do clip the vowel, However, seemingly, English pre-fortis clipping is more pronounced than this physiological effect can explain, and Serbian clipping falls far short of this. This is particularly noticeable for diphthongs and long vowels, and as a consequence, pairs such as ‘loose’ and ‘lose’ or ‘tied’ and tight’ become indistinguishable.

Weak Syllables

The most common sound in English is the reduced vowel schwa, and it is very important in creating the prosody that characterises the language.

Kristyna Poesová and Lenka Weingartová pointed out that in a way, the importance of the reduced vowel is negative: they highlight the vowels which aren’t reduced. This contrast is lost if you fail to reduce vowels sufficiently. They chose to study the acquisition of schwa among Czech speakers of English. Seemingly, while Czech learners have some success in picking up schwa in the unstressed syllables of longer words, and before an R in words like doctor, they rarely use it in weak forms (except in the case of a and the), so this requires some careful instruction.

Jan Volin and Terezie Johaniková focused on weak forms because of their importance in the rhythm of English. Their study showed that Czech speakers tend to strengthen weak forms that begin with a vowel sound such as ‘at’ or ‘of’ by preceding them with  glottal stop. Regarding elision, they found that while Czech speakers may acquire such elisions of D in and, they rarely go for the more radical elisions that characterize spontaneous spoken English such as the R in from.

Spontaneous Discourse

Beatrice Szczepek Reed used her plenary slot to argue the case for looking at pronunciation in the context of discourse, taking a conversation analysis (CA) approach. She illustrated the idea with a close focus on the strategic use of glottal stops in German and English. The default position in these two languages couldn’t be more different – while in German, words which begin with a vowel are preceded by a glottal stop, in English they are linked to the preceding word. However, in both languages, the speaker has the choice to break away from this default position. In German, a speaker may refrain from using a glottal stop in order to slip in a new proposition without too much of a fanfare. For example, a criticism can be promptly neutralized by slipping in a softener not preceded by a glottal. English speakers use the opposite strategy – they can create more of a fanfare for a new proposition by emphasizing it with a glottal stop.

Richard Cauldwell used his all-too-brief slot to argue that pronunciation models for listening must be different from speaking. The rules and patterns described in most pronunciation models maybe useful, but are patently false as descriptions of how people actually speak. He presented an analogy of greenhouse (citation forms of words), garden (connected speech following the usual rules presented in pronunciation books) and jungle (the way words are blended together and reduced in actual spontaneous speech).

John Levis and Greta Muller Levis reported a study showing how speakers signal layers in their speech, with parenthetical asides being signalled by reduced prominence. The communicative value of non-prominence in a lecture scenario, for example may be something like: ‘you don’t need to take a note of this bit’.

Models and Targets

Ewa Waniek-Klimczak, in her plenary slot, argued for making a distinction between learners and users of English, and made the point that the same individual may be a learner in the classroom but a user on the street. The notion of ‘correct’ is only appropriate for the former. She then went on to distinguish models from targets. The model may be a standard form of English such as RP. This model will consist of a set of meaningful contrasts which the learner needs to be able to distinguish, but they may do this in their own L1 accent. So: L2 phonemes with L1 allophones (my gloss). For a monolingual group, we can set appropriate targets by analysing their L1 and working out which contrasts they already have in their own language and which will need to be learnt.

Rias van den Doel also considered the question of models. He criticized the idea of one global lingua franca core, arguing that this would be just as hegemonic as the RP or other prestige variety that it replaced. He also argued against setting local targets such as Polish English or Spanish English, saying that in the given country, this overlooks variety and minorities and becomes prescriptive again. Such local endonormative targets are as disempowering as RP.

Radek Skarnitzl and Alice Henderson talked about the unfortunate reality of prejudice against foreign-accented speakers, who are seen as being less reliable because of their accent – and ironically, the harshest judges are those with the same L1. And the implication? Do we need to train away the prejudice?

Dan Frost and Jean O’Donnell used their short slot to introduce us to their CEFR-style rating scales for prosody – a monumental piece of work in progress. These descriptors will raise teacher awareness of what needs to be taught. It will also enable better assessment and create washback, so that pronunciation will no longer be neglected as an optional add-on. This will benefit speaking, but even more importantly, listening.

Pronunciation Pedagogy

Pekka Lintunen presented a study revealing the paradox that the more pronunciation training you get, the less confident a speaker you are. It seems that prior to training, we may be in a state of blissful ignorance, not knowing what we don’t know. But perhaps things have to get worse before they can get better!

Celine Horgues and Sylwia Sheuer presented videos showing pairs of English-speaking and French interlocutors conversing in either English or French. We focused on examples of the native speaker in each pair giving correction. It seems that this was mostly done indirectly, as recasting. The speakers were in a non-hierarchical relationship and so avoided being too controlling or didactic. The focus of the correction was rarely pronunciation, and where it was it was, it was usually segmental features. The correction was often overlooked by the recipient because the recast was not recognized as being a correction.

Małgorzata Baran-Łucarz examined the importance of ‘willingness to communicate’. Language has a dramatic effect on this, so that a person who is gregarious in L1 may be very reticent in L2, and this would depend on their ‘Self-perceived communcation competence’ (SPCC). It seems that while poor grammar does not damage SPCC much, pronunciation definitely does, and especially in the classroom context. This is often exacerbated because the classroom is seen as a place of evaluation rather than learning, so learners are afraid to stick their necks out and make mistakes.

Marta Nowacka examined the potential of videos from the internet for raising awareness of aspects of pronunciation in a fun way. She showed examples including adverts and short excerpts.

My own plenary presentation looked at classroom techniques and materials relating to individual sounds – more information here.