Robin Walker on technology in pronunciation teaching

Robin Walker (read our review of his latest book here) began by positing 3 stages in acquiring pronunciation: 1. the cognitive stage - becoming aware of a feature; 2. the associative stage - training yourself to be able to deal with the feature; 3. the autonomous stage - becoming able to do it automatically and fluently. Good pronunciation lessons - and by extension, tech tools - need to contribute to all these stages. For this, they need to explain and justify the need for a particular focus, give intensive practice, and provide meaningful feedback and progress checking.

Early CALL materials provided very few of the above. For example, you might hear one word of a minimal pair and be expected to identify which one it is. There was no demonstration of how the sounds are produced, why they're meaningful, or what to do if you are UNable to identify the correct answer. And the above setbacks are also true of many currently promoted digital material, including digital add-ons to otherwise excellent printed materials.

Regarding the need to justify the need for a given pron focus, many tech tools fail in their hidden assumption that the learner needs to learn to mimic a given prestige native variety of English. For instance, while voice recognition technology can be used by students to work on their pronunciation by trial and error, they will also waste time correcting features of their pronunciation which are not erroneous, merely NOT what the computer was programmed to expect. Robin himself was misunderstood and corrected by speech recognition tech. And he illustrated the problem with the Youtube sketch about two Scots trapped in a speech recognition lift.

The actual tools that Robin presented were classified into 1. listening, 2. tuition and 3. recording. Regarding listening, Robin's top tips were elllo.org for accent variety and Richard Cauldwell's Cool Speech app for well explained, focused micro-listening.

For tuition, there was no unqualified recommendation (but several duds, by the criteria outlined above. The most promising-sounding material was the University of Iowa site. Finally, regarding recording tools, Robin recommended Wave Pad as more suitable than the oft-cited Audacity - this latter being like taking a sledgehammer to open a walnut for most teachers' purposes. Robin also recommended the voice recognition app Dragon, which is capable, after a period of training, of adapting to the teacher's own accent. As Robin said, Dragon proves to be incapable of accurately interpreting weak forms, but for international English, this is not required - speakers will be more globally intelligible if they DON'T use weak forms.

On our map of ELT, I would locate this talk on the frontier between North East (learner needs) and the South East (Teaching and learning tools), fairly close to the centre.