The attitudinal function of intonation

Many writers have expressed the view that intonation is used to convey our feelings and attitudes: for example, the same sentence can be said in different ways, which might be labelled "angry", "happy", "grateful", "bored", and so on. A major factor in this is the tone used, and most books agree on some basic meanings of tones. Here are some examples (without punctuation):

1 Fall

Finality, definiteness: That is the end of the ews

                                       Im absolutely certain

                                       Stop alking

2 Rise

Most of the functions attributed to rises are nearer to grammatical than attitudinal, as in the first three examples given below; they are included here mainly to give a fuller picture of intonational function.

General questions:               Can you /help me

                                                Is it / over

Listing:                                    /Red /brown /yellow or lue

                                                (a fall is usual on the last item)

"More to follow":                 I phoned them right a/way ('and they agreed to come')

                                                You must write it a/gain (and this time, get it right)

Encouraging:                         It wont /hurt

3 Fall-rise

Uncertainty, doubt:             You vmay be right

                                                Its vpossible

Requesting:                          Can I vbuy it

                                                Will you vlend it to me

4 Rise-fall

Surprise, being impressed: You were a first

                                               vAll of them

It has also been widely observed that the form of intonation is different in different languages; for example, the intonation of languages such as Swedish, Italian or Hindi is instantly recognizable as being different from that of English. Not surprisingly, it has often been said that foreign learners of English need to learn English intonation. Some writers have gone further than this and claimed that, unless the foreign learner learns the appropriate way to use intonation in a given situation, there is a risk that he or she may unintentionally give offence; for example, the learner might use an intonation suitable for expressing boredom or discontent when what is needed is an expression of gratitude or affection. This misleading view of intonation must have caused unnecessary anxiety to many learners of the language.

Let us begin by considering how one might analyze the attitudinal function of intonation. One possibility would be for the analyst to invent a large number of sentences and to try saying them with different intonation patterns (i.e. different combinations of head and tone), noting what attitude was supposed to correspond to the intonation in each case; of course, the results are then very subjective, and based on an artificial performance that has little resemblance to conversational speech. Alternatively, the analyst could say these different sentences to a group of listeners and ask them all to write down what attitudes they thought were being expressed; however, we have a vast range of adjectives available for labelling attitudes and the members of the group would probably produce a very large number of such adjectives, leaving the analyst with the problem of deciding whether pairs such as "pompous" and "stuck-up", or "obsequious" and "sycophantic" were synonyms or represented different attitudes. To overcome this difficulty, one could ask the members of the group to choose among a small number of adjectives (or "labels") given by the analyst; the results would then inevitably be easier to quantify (i.e. the job of counting the different responses would be simpler) but the results would no longer represent the listeners' free choices of label. An alternative procedure would be to ask a lot of speakers to say a list of sentences in different ways according to labels provided by the analyst, and see what intonational features are found in common - for example, one might count how many speakers used a low head in saying something in a "hostile" way. The results of such experiments are usually very variable and difficult to interpret, not least because the range of acting talent in a randomly selected group is considerable.

A much more useful and realistic approach is to study recordings of different speakers' natural, spontaneous speech and try to make generalizations about attitudes and intonation on this basis. Many problems remain, however. In the method described previously, the analyst tries to select sentences (or passages of some other size) whose meaning is fairly "neutral" from the emotional point of view, and will tend to avoid material such as 'Why don't you leave me alone?' or 'How can I ever thank you enough?' because the lexical meaning of the words used already makes the speaker's attitude pretty clear, whereas sentences such as 'She's going to buy it tomorrow' or 'The paper has fallen under the table' are less likely to prejudice the listener. The choice of material is much less free for someone studying natural speech. Nevertheless, if we are ever to make new discoveries about intonation, it will be as a result of studying what people really say rather than inventing examples of what they might say.

The notion of "expressing an emotion or attitude" is itself a more complex one than is generally realized. First, an emotion may be expressed involuntarily or voluntarily; if I say something in a "happy" way, this may be because I feel happy, or because I want to convey to you the impression that I am happy. Second, an attitude that is expressed could be an attitude towards the listener (e.g. if I say something in a "friendly" way), towards what is being said (e.g. if I say something in a "sceptical" or "dubious" way) or towards some external event or situation (e.g. "regretful" or "disapproving").

However, one point is much more important and fundamental than all the problems discussed above. To understand this point you should imagine (or even actually perform) your pronunciation of a sentence in a number of different ways: for example, if the sentence was 'I want to buy a new car' and you were to say it in the following ways: "pleading", "angry", "sad", "happy", "proud", it is certain that at least some of your performances will be different from some others, but it is also certain that the technique for analyzing and transcribing intonation introduced earlier in the course will be found inadequate to represent the different things you do. You will have used variations in loudness and speed, for example; almost certainly you will have used different voice qualities for different attitudes. You may have used your pitch range in different ways: your pitch movements may have taken place within quite a narrow range (narrow pitch range) or using the full range between high and low (wide pitch range); if you did not use wide pitch range, you may have used different keys: high key (using the upper part of your pitch range), mid key (using the middle part of the range) or low key (the lower part). It is very likely that you will have used different facial expressions, and even gestures and body movements. These factors are all of great importance in conveying attitudes and emotions, yet the traditional handbooks on English pronunciation have almost completely ignored them.

If we accept the importance of these factors it becomes necessary to consider how they are related to intonation, and what intonation itself consists of. We can isolate three distinct types of suprasegmental variable: sequential, prosodic and paralinguistic.