A couple of weeks ago, I woke up with a nasty cold.
A couple of days later, I woke up to discover I’d lost my voice.
A couple of hours after that, I was due to teach a CertTESOL input session.
Cue Mission:Impossible music…
Well, they say necessity is the mother of invention, so I threw my notes from the ‘normal’ version of the input session out the window (not literally – my school has surprisingly few windows) and got inventing!
As I believe one of the best ways to help trainee teachers (and language learners, for that matter) understand pronunciation is to get them to make sounds themselves and listen to each other (as in, not just listen to me ‘explain’), and since I didn’t really have any other option in this instance, not having the facilities to subtitle myself, that’s what we did. With a little help from PowerPoint.
Today’s focus: connected speech. In the previous two input sessions, the trainees had already studied the phonemic chart and word stress. So they were equipped with the basics for stringing sounds together and discovering what can happen (at least, in some accents of English).
Er, see above.
Or “How To Teach Connected Speech In 5 Simple Steps, Without Actually Uttering A Word”.
Note: throughout the following procedure I was 100% silent, except for occasional coughing.
1. I prepared several short sentences in PowerPoint which described my predicament, transcribed phonemically. I had to think for a while about how to exemplify various features without my sentences sounding really forced, and I carefully avoided in this transcription any features of connected speech (my reasons will become clear later).
2. I revealed these one by one on the board, allowing the trainees to gradually read them and sound them out, then tell me what they said.
Here’s the whole text as it looked once all the sentences had been revealed and deciphered:
Transcription, if you need it:
I’ve lost my voice.
Since I can’t speak,
you’ll have to.
Do all you can do to help each other.
Before asking me to help, see if your partner can.
3. So far, so good. Now we needed to find out what these sentences would really sound like, if spoken in RP*. I needed to make this distinction (between unconnected, lego-style piecing together of phonemes, and rapid, connected RP) very early on as, in my experience, this avoids huge arguments throughout the session about how something might (not) sound ‘in reality’. So I revealed and highlighted (by underlining repeatedly with a board pen and gesticulating madly and mutely) the following colour-coded key in the top right corner:
4. Then I repeated what I did in step 1, but this time revealing one by one the connected speech version of each sentence. For example:
The trainees very quickly grasped that what they had initially looked at was essentially a list of words, transcribed individually; but that they were now looking at how these words sound when strung together. (Admittedly, a significant amount of mad gesticulation and jabbing at the key in the top right-hand corner helped them understand this.)
As I revealed each line, I also revealed a red box which highlighted the particular connected speech feature evident in that utterance, for example:
The trainees’ job at this point was to “ooh” and “ahh” and work out what had happened. Fortunately, for each and every feature I revealed, there was at least one trainee in the group who’d encountered this before and could explain it more thoroughly to the others, if they were unclear/unsure on what was happening.
(Aside: I love it when this happens in a class. It’s soooo nice to see trainees/students really listen to each other so attentively and help each other so earnestly, and it makes my job so much easier as I can see much more clearly where there are gaps in their knowledge, then just try to guide them to plug these.)
And here’s how the whole board looked once everything had been revealed:
As each feature was revealed with a red box, then identified and explained by the trainees themselves, I managed to silently elicit the names for these features. There were only a couple that they didn’t know. I annotated these in red under the boxes with my board pen as we went through them all.
I also elicited explanations of these features from the trainees (e.g. that intrusive /j/ and /w/ occur based on which sounds are neighbours over a word/syllable boundary) and added these notes on the board, for the trainees’ reference.
In fact, by the time we’d done the first 5 lines, they were all doing very well, even those trainees who weren’t feeling very confident about phonology. So before I revealed the last line, I got them to discuss in pairs what features they’d expect to see. They did exceptionally well at predicting, identifying and naming these, and I was feeling quite proud of them (and myself) at this point. (Though not as proud as I felt a few days later, when most of them performed better than I’m used to seeing on the connected speech part of their phonology test!)
5. Introducing this theory in this input session normally takes about 60 minutes (and often more, what with all the arguing back and forth about what people ‘really’ say), leaving very little time for a worksheet that provides further practice in identifying these features in other examples of extended speech. But this time, we got through all of that in only 30 minutes AND the trainees seemed to have a much better understanding than usual. So there was plenty of time for further practice, and there were very few problems or questions.
You can download the slides I used here, if you ever find yourself in the same situation (or just want to trick your trainees into teaching themselves connected speech).
*As the standard phonemic chart used in Europe is based on RP, we took that as the benchmark for our transcriptions in this session. No implications were intended as to ‘correctness’, but we just needed to establish very early on that some of the features we were going to meet would not be relevant to other accents, so trainees should suspend their disbelief and just accept that we were working within a certain framework for now, in order to help them grasp some potentially confusing linguistic theory.