Intonation: Guidelines, not rules

by | 22 February 2023

When experienced actors can’t deliver the lines, you know you have a problem with the script. This was our issue with the first videos for the Clear Pronunciation 2 Intonation section, filmed over a year ago. The intonation produced by the voice talents simply wasn’t what we had expected. Worse, when they were prompted to say the model sentences ‘correctly’, they sounded stilted and unnatural.

This wasn’t due to shortcomings in the actors – they are specialists and had already completed the word and sentence stress, connected speech and consonant clusters sections with no difficulty at all. So why couldn’t they do the same with the intonation models? Answering this question goes beyond improving the videos. The direction we give to the voice talents pretty much mirrors the instructions and modelling teachers give to their students. So it’s worth doing a bit of digging.

Let’s start by agreeing on the basic ‘rules’ for intonation. These are neatly summarised by the British Council:
– Wh-word questions: falling intonation
– Yes/No questions: rising
– Statements: falling
– Question-Tags: ‘chat’ – falling; ‘check’ – rising
– Lists: rising, rising, rising, falling

Perhaps counter-intuitively, the same basic patterns apply not just to British English, but also to General American and General Australian. (Here’s an Introduction to Intonation from City University New York.)

That all seems fairly clear, but let’s look at an example of where it goes wrong. Following one of the guidelines above, we tell students that in Yes / No questions, the intonation rises ‘at the end of the question’. This is confusing because, in reality, the tone of the voice is equally likely to fall at the end of the question. Consider these two sentences:

  1. Are you seeing her soon?
  2. Are you seeing her tomorrow?

In the first question, the tonic syllable, where the pitch change occurs, is ‘soon’, the final syllable of the sentence, so the intonation does indeed rise at the end. But in the second sentence the pitch rises on the penultimate syllable of the final word ‘tomorrow’, and actually falls on the final syllable. It’s therefore essential that students understand the concept of the tonic syllable, and learn to identify it. That’s the only way they can interpret ‘at the end of the question’ correctly.

Next, we need to consider how attitude and intent affect intonation. It’s not difficult to think of a context in which the intonation in either of the questions above would fall at the end rather than rising. Simply picture an angry parent asking the second question:
‘Let’s get this straight. Are you seeing her tomorrow?’

The intonation falls on ‘tomorrow’; try saying it with rising intonation, and it sounds very peculiar. We (the ESOL profession) have decided that a ‘neutral intent’ is the default and to base the intonation patterns we teach on that. That is something of an arbitrary decision, and we therefore need to ensure that students understand that we are dealing not with absolute rules, but with shades of grey. This is not the case with other areas of pronunciation. In word stress, everything is black and white: we always put the stress in the word ‘crocodile’ on the first syllable.

And there are other elements we need to factor in. Take this exchange (as we did in our video session):
‘Jack’s coming tomorrow.’
Even contextualising it by saying that the second speaker is surprised is not enough. We also need to know how the second speaker feels about Jack . If they like him, the intonation might fall; if they don’t, it might rise. How strongly do they feel? That might determine both the volume, and the amplitude of the pitch change. And all these are ‘mights’. So even when modelling the exchange with a native speaker, the variables place such a massive cognitive load that it is impossible to predict the output. (In fact, about once a month I catch myself coming out with this exact surprised utterance: ‘Tomorrow?’ Each time I try to analyse my intonation, and each time I find it terribly difficult.)

Finally, we need to consider the influence of the learner’s L1. In Chinese languages, pitch is used to distinguish the meaning of one word from another. In French the pitch range is relatively small; in Malay it is extremely broad. If you’d like to explore further, search for ‘Intonation Systems: A Survey of Twenty Languages’ where you can download a 505-page CUP book on the subject. (That speaks for itself.)

So, where does that leave us? It’s fair to say, as the British Council does, that the descriptions above are ‘starting-points, rather than rules’. It’s useful to give students these guidelines, provided we ensure they understand that’s what they are. But given the significant challenges, perhaps with intonation it’s more realistic to focus primarily on listening rather than speaking: to help students interpret the intonation of other speakers of English, particularly native speakers, rather than insisting on them producing it accurately themselves. Intonation rarely causes a breakdown in intelligibility – another reason why it’s good to focus more effort on recognising what you hear than worrying about trying to produce pitch changes which are subtle and nuanced.

We ended up changing the emphasis of the intonation section along these lines and re-shooting all the intonation videos. This iteration feels both more tentative, and more useful. Take a look at the Introduction video (British English version) here:

You can learn more about Clear Pronunciation 2 here.

Andrew Stokes, Publisher, ClarityEnglish

Andrew Stokes, Publisher, ClarityEnglish