The definition of a tone unit is a “a stretch of speech uttered under a single … intonational contour” (Du Bois 1992). The key to this definition is a ‘single intonational contour’. Thus the decision as to where to split a string of words based on tone units depends primarily on the change in tone. Pausing and lengthening of syllables can also be used as cues for the termination of the unit although these are secondary.
Thus in the following snippet of speech uttered by David Cameron in the House of Commons during Prime Minister’s questions, there are three tone units each with its own intonational contour.
DC: well I’ve given a very \fu:ll answer to this \obviously (.5) I regret (.3) the decision to employ Andy Coulson on the basis (.) of the assurances that I was \given
The tone units below, demarcated by the | vertical bar | character, have been determined primarily on change in tone rather than pausing. If pausing was the main criteria then we probably would not separate units 1 and 2 because there is no significant pause between them. There is however a significant pause between units 2 and 3.
- | well I’ve given a very full answer to this |
- | obviously |
- | I regret the decision to employ Andy Coulson on the basis of … |
This then begs the question as to what the speaker (David Cameron) meant ‘obviously’ to attach to: the preceding tone unit or the following. Did he mean that he had obviously given a ‘full answer’ or did he mean that he obviously ‘regret[s] the decision’? The official Hansard transcription, published by the Houses of Parliament, has ‘obviously’ attaching to the following tone unit:
The Prime Minister: I have given a very full answer to this. Obviously, I regret the decision to employ Andy Coulson on the basis of the assurances I was given (Hansard)
However, there is a degree of ambiguity in how it should be interpreted and the pausing suggests the preceding unit. The intonation trace given below suggests that ‘obviously’ is closer to the preceding unit that the one that follows and a person listening at the time might have assumed that this is the case.
Fortunately cases like these are not so common and in generally the demarcation of tone units is a relatively straight forward affair.
For more on the tone-unit and how it is defined, see the post here.
Du Bois, J. W., Cumming, S., Schuetze-Coburn, S. & Paolino, D. (1992), Discourse Transcription, Santa Barbara Papers in Linguistics, 4, Santa Barbara, CA, Department of Linguistics, University of California, Santa Barbara.