Listening Cherry 13 – Connected speech rules are too genteel

Image from here

In my work I distinguish between three styles of speech: the Greenhouse, the Garden, and the Jungle. The Greenhouse is the domain of the citation form, where each word is presented in isolation, with all its features perfectly represented, un-interfered with by other words. The Garden is the domain of the rules of connected speech, where words are in orderly and pleasing arrangements and where they glide into each other, with genteel touches (handshakes) and make slight changes in sound shapes at their boundaries. Words behave politely, in a way that appropriate for those genteel occasions when you are having tea on the lawn (‘Would you like another cup of tea dear?’ becomes ‘Wu jew lie ka cuppa tea dear?). The Greenhouse and the Garden are useful for teaching pronunciation, and clear intelligible speech. The Jungle is real life speech, where words are mangled, crushed, bashed in a disorderly mess – speed and lack of clarity are the order of the day (‘July annuvver cuffer tea pop?’). The Jungle is where we need to go if we are to improve the teaching of listening.

Our much cherished rules of connected speech belong in the garden. My contention is that they are therefore too genteel, which I now hope to demonstrate.

The following example comes from a textbook that I have used recently. It illustrates the linking sounds /r/, /w/ and /j/ in the underlined parts:

These are enormous sums of money 
for people to actually invest in in cough cold remedies (Oakey & Treece, n.d.)

We look at such examples, and inspect them at leisure, and sound out to ourselves the genteel handshakes between the words which are joined by the linking sounds.

• There is an /r/ linking ‘are’ and ‘enormous’ – thus ‘ah-ree-normous’
• There is a  /w/ linking ‘to’ and ‘actually’ – thus ‘too-wack-shuh-ly’
• There is a  /j/ linking ‘actually’ and ‘invest’ – thus ‘ak-chew-lee-yin-vest’

This is a very genteel activity – a slow-paced, good-mannered approach to word contact in which words are savoured at leisure. But when you listen to the recorded extract below (from a lecture) these linking phenomena are far from salient, there is so much more going on in the sound substance, including very ungenteel word-bashing and squeezing. These other phenomena drown out our linking sounds, and they are very difficult to perceive, and they may even be absent. Have a listen:

These are enormous sums of money 
for people to actually invest in in cough cold remedies

(Oakey & Treece, n.d.)

There are sixteen words (counting both occurrences of ‘in’) and twenty-five syllables (counting three in ‘actually’) going at an average of 5.7 syllables per second – which is fast, much faster than is likely in our leisurely savouring of the linking.

For an example of word-bashing, listen to the words ‘enormous sums’. You should hear that the final syllable of ‘enormous’ does not occur, it disappears into the initial segment of the following word ‘sums’ – as the result of a very non-genteel bashing. So we get

‘enorm sums’.

A transcription reveals other things about this extract.

01 || these are eNORMous || 6.9

02 || SUMS of MONey || 7.7

03 || for PEOPle to actually inVEST in || 5.8

04 || in COUGH cold REmedies || 3.7

These are four speech units (transcribed following the conventions of Cauldwell, 2013; Brazil, 1997). Each line is a speech unit, upper case letters denote prominent syllables, lower case letters denote non-prominent syllables; the ‘linking’ is shown as underlining. The speed of each speech unit is given in syllable per second at the end of teach speech unit. Each speech unit has a different speed, with a high of 7.7 and a low of 3.7 syllables per second. Notice that the predicted linking sounds occur in non-prominent syllables, which are going really fast in the squeeze zones (Cauldwell, 2013, Chapters 2/18) of the speech units.

And if we consider the linking in 01, there is a case to be made for the linking /r/ not being present at all. Listen to the following extract which gives the words ‘are enormous sums’ at both full speed and half speed. Because we know that they contain the words ‘are enormous’, we are primed to hear them (see the work of Helen Fraser for more on priming).

are enormous sums/’ah ee nawm sums'[ainɔms].

It seems to me an equal possibility that the sound substance has no [r] at all. Listen to the extract again, and pretend the speaker is saying the word ‘ginormous’, but without the first segment, and without the last syllable – thus ‘-inorms’ [ainɔms]. If you do this, you may not hear the ‘linking /r/’ at all. When I listen in this way, I hear an [a] from ‘are’, plus [i] from ‘enormous’ giving a sound something like the first person pronoun, and emphatically not  ‘ah-ree-normous’.

There is so much more going on in spontaneous speech than is allowed for in our ELT textbooks. Exercises involving the slow enacting of linking sounds are too tidy, too genteel – ‘too garden’ – for the purposes of teaching listening. We need a description of spontaneous speech to help us and our students in our encounters with the Jungle.
Brazil, D. (1997). The Communicative value of intonation in English. Cambridge: Cambridge University Press.Cauldwell, R.T. (2013). Phonology for listening. Birmingham: Speech in Action.
Oakey, D. & Treece, P. (n.d.) English for academic purposes programme: Developing listening skills & giving presentations. EISU, University of Birmingham.