A demonstration of David Brazil’s theory of Discourse Intonation

Table of contents

1. Introduction

2. Brazil’s Discourse Intonation - Outline
2.1 The tone unit
2.2 The tone
2.3 Key and termination

3. Intonational analysis
3.1 Demonstration ofBrazil’s Discourse Intonation - Procedure
3.2 Demonstration ofBrazil’s discourse intonation using the example of a short extract of The Lord OfThe Rings

4. Conclusion

5. References

6. Appendix

Introduction

This paper will deal with the analysis of short linguistic units on the basis of David Brazil’s theory of discourse intonation. In the first part of the work presented here we will therefore give a general outline of Brazil’s theory and his systematic approach to intonation analysis. In the second part we will proceed to the main subject of the paper, that is, we will examine a short text for potential intonations.

This paper is aimed neither at presenting a new linguistic approach nor at critically analysing an existing theory but rather at exemplifying such a theory. We will focus on Brazil’s discourse intonation as it is presented in his book Pronunciation for Advanced Learners of English.

1. Brazil’s Discourse Intonation

Intonation constitutes an essential element of language in oral communication and is thus subject of linguistic research. While the assumption that intonation carries meaning goes widely unchallenged, linguists still argue as to the exact function of intonation.

David Brazil, along with other linguists, propagated the idea of intonation as being discoursal in function. He thus distances himself from those who conceive intonation as having a grammatical, accentual or attitudinal function.

1.1. The tone unit

According to Brazil, intonation “marks the turn-taking processes in an exchange between speakers”¹, i.e. the way a speaker varies the pitch and the prominence over a stretch of speech mainly depends on what he/her assume to be “the state of understanding between him/her and the hearer”².

Based on this position, Brazil’s discourse intonation distinguishes between four systems of speaker choice: prominence, key, termination and tone. These systems develop within the framework of so-called tone units. Tone units, or tone groups, are “basic building blocks of spoken English.”³ Rather than viewing spoken language as a sequence of separate words, Brazil sees it as being made up of groups of words the sounds of which “are usually run together in the way we are accustomed to thinking of the separate sounds of single words as being run together.”⁴ In this construct of ideas, a tone unit may contain either one or two prominent syllables, the last one of which is always tonic by definition. The assignment of prominence to any word, irrespective of whether it is also tonic or not, “depends upon whether the word is selective or not in the particular circumstance of the utterance.”⁵ That is, if a word is used to remove uncertainties and was thus purposely chosen from a set of possible other words, it is made prominent. The essential difference that distinguishes a tonic syllable from a prominent, non-tonic syllable, is the fact that the first has a characteristic change of pitch. These changes of pitch are subdivided into different types of tones which we will specify in chapter 1.2.

The criteria Brazil established in order to define a tone unit are quite controversial. It has often been implied that Brazil’s fragmentation of utterances is not representative of the natural stream of speech. Criticism is particularly directed at Brazil’s insistence on the fact that a tone unit can never contain more than two prominent syllables. In a review article on the book “Discourse Intonation and Language Teaching” Windsor Lewis said about Brazil’s theory:

“Boundaries [of tone units] are shown in numbers of cases where the rhythmic integration of the words on either side of the double slash is maximal [i.e. a two-prominence tone unit], for example on the several occasions which assign the weak article the to a separate unit from its following substantive.”⁶

The fact that discourse intonation “does not attribute any significance to the location of boundaries”⁷ and consequently does not give an exact definition as to where one tone unit ends and another begins seems to directly contradict the assumption that the sounds of a tone group are run together. Brazil confines himself to saying that “a tone-unit ends somewhere between the occurrence of a tone, and the onset prominence of the following tone-unit.”⁸

Each tone unit represents a separate parcel of information so that each parcel of information, if interpreted as such by the speaker, requires a new tone unite. This is a very vague description because what is and what is not considered as such a parcel is up to the speaker and can not always be deduced from the text itself. That is why, to a certain extend, a written text leaves it to the reader to decide which information belong to one parcel and which ones belong to another and thus which syllable is to be made tonic and which syllable is to be made prominent but not tonic. This ambiguity, however, prompted critics, such as Lewis, to accuse the theory of being arbitrary:

“At any rate their [i.e. Brazil et al.] claim to have established a 'general basis for distinguishing prominent syllables from tonic syllables' is quite unjustified and their transcriptions show large numbers of cases where their choice between showing a tone as onset or tonic appears to be arbitrary.”⁹

These problems have to be taken into consideration when generating speech on the basis ofBrazil’s theory of discourse intonation.

1.2 The tone

As we have already said, “the tonic syllable is the syllable on which the main pitch movement begins.”¹⁰ It constitutes the core of a tone unit and thus determines the tone of a tone unit. The tone, in turn, is the carrier of intonational meaning. That is to say, meaning that is conveyed by means of intonation evolves from the tonic segment of a tone unit.

Phoneticians have classified five typical pitch movements or tones, namely the fall, the fall-rise, the rise, the rise-fall and the level tones. According to Brazil, each of these intonation patterns has its own function. They are either referring tones (rise and fall-rise tones) or proclaiming tones (fall and rise-fall), that is to say, they are used either in contexts where the conveyed information is considered by the speaker as being already known to the hearer or in contexts where the message is considered to be unknown to the hearer. The level tone is an exception within this system of tones. It is neither proclaiming nor referring but “it signifies that the speaker has a focus on the wording which he/she is compiling, rather than on interpersonal interactivity.”¹¹

A further distinction has to be made between rise and fall-rise tones on the one hand and between fall and rise-fall tones on the other. Unlike the fall-rise, the rise tone is often used in situations where the speaker wants to emphasise the fact that he/she occupies the role of the dominant speaker. In most other cases where referring tone has to be adopted, the fall-rise is preferred.

In terms of the two proclaiming tones, the fall indicates that the (knowing) speaker provides the hearer with new information while the rise-fall is used when the information is new to both speaker as well as hearer. Thus, the rise-fall occurs where the speaker wants to express surprise, irritation and strong personal impression. It is, however, very rarely to be found in actual speech, which is probably the reason why, in his book Pronunciation for Advanced Learners of English, Brazil failed to give further explanations as to how and when exactly this tone is used.

Having explained the meaning of prominence and tone in speech, we will now proceed to focus on the other two aspects of intonation, namely key and termination.

1.3. Key and termination.

In the context of intonation, the term key refers to the pitch level of a given onset syllable while termination denotes the specific pitch levels of tonic syllables. Depending on the relative pitch level of a given prominent or tonic syllable, key and termination have to be classified as being either low, mid or high.

Each pitch level represents a specific intonational meaning. High key is used when the “matter of the tone unit is presented as being contrary to the expectations of the hearer”¹² whereas low key occurs where the message is supposed to perfectly meet the hearer’s expectations. Mid level indicates that there is no particular assumption upon the part of the speaker about what the hearer’s expectations are, that is, the content of the respective mid key tone unit isjust an addition to what has already been said before.

The function of termination differs from that of key in that it is not so much concerned with the hearer’s expectations but rather with the speaker’s expectation with regards to the key of the hearer’s response. Thus, high termination signals that the speaker expects a high key response from the hearer, e.g. adjudication, while mid termination anticipates a mid key response as in concurrence. Low determination has the special function of closing a pitch sequence. ¹³

2. Intonational analysis

The following chapter will be concerned with the illustration of the theory of discourse intonation. We will use a short extract from J. R. R. Tolkien’s The Lord of the Rings as an example.

We will start by explaining briefly how we are planning to proceed.

2.1 Illustration of Brazil’s discourse intonation - Procedure

Since, in terms of intonation, the full stop is a natural tone boundary (pitch sequence), we will analyse the text sentence by sentence. We will start by retelling briefly what the sentence is about in order to identify the selective slots in it and to explain why we defined one word as occupying a selective slot or giving new any information while we did not do so with another one. Having thus spotted the prominent syllables, we will decide which of those prominent syllables have to be made tonic, that is, which of those syllables carry intonational meaning. At the same time we will have to indicate tone unit boundaries since it is impossible to recognize a tonic syllable without knowing which parts or groups of words of the respective pitch sequence are intended to form separate parcels of information and which ones are supposed to belong together.

Lastly, we will decide if any statements regarding the choice in the three-term system of pitch level can be made.

Following David Brazil’s example¹⁴, we will mark prominent syllables by capitalising them and tonic syllables by capitalising and underlining them. Tone unit boundaries will be signalled by pairs of slashes and the tones will be indicated by arrows (see table below) immediately after the initial boundary sign.

As to the pitch level, if prominent syllables are supposed to be placed higher or lower compared to the previous prominence, they will be preceded by up or down arrows.

illustration not visible in this excerpt

In order to avoid unnecessary repetition, we will only explain in detail the transcriptions of the first half of the text and confine ourselves to merely giving the transcriptions for the second half of the text.

[...]

¹ Skandera, Burleigh 2005, p. 119

² http://www.speechinaction.net/SPARC DI.htm (2008-02-08)

³ Brazil 1994, p. 7

⁴ ibidem

⁵ Brazil 1995, p. 90-100

⁶ Windsor 1986

⁷ http://www.speechinaction.net/SPARC DI.htm (2008-02-08)

⁸ ibidem

⁹ Windsor 1986

¹⁰ Skandera and Burleigh 2005, p. 121

¹¹ http://www.speechinaction.net/SPARC DI.htm (2008-02-08)

¹² Brazil 1995, pp. 244-246

¹³ “A pitch sequence is a stretch of speech consisting of one or more tone units and ending with low termination”, Brazil 1995, p. 246

¹⁴ see e.g. Brazil 1994; in older books (e.g. BRAZIL, David: Discourse Intonation andLanguage Teaching. O. Longman, London 1980) Brazil used different symbols