Underlying any talk of 'rhythm' is the notion of a regular periodic beat; and the very fact that we apply this term to language means that some analogy is drawn between a property of language, and the ticking of a clock, the beat of a heart, the step of a walker, and other regularly recurrent happenings in time. In phonological discussion, the grandiose term isochronism ('equal-time-ness') is attached to this simple principle. To attribute the isochronic principle to a language is to suppose that on some level of analysis, an utterance in that language can be split into segments which are in some sense of equal duration. In certain languages, such as French, this segment is the syllable. In others, such as English, it is a unit which is usually larger than the syllable, and which contains one stressed syllable, marking the recurrent beat, and optionally, a number of unstressed syllables. This is the unit that I have previously called the (rhythmic) measure. Thus English and French are representatives of two classes of language, the 'stress-timed' and the 'syllable-timed' respectively.4
I have emphasized the qualification 'in some sense of equal duration', because the rhythm of language is not isochronic in terms of crude physical measurement. Rather, the equality is psychological, and lies in the way in which the ear interprets the recurrence of stress in connected speech. Here there is a helpful analogy between speech and music. A piece of music is never performed in public with the mechanical rhythm of the metronome, and yet despite various variations in tempo, some obvious and deliberate, some scarcely perceptible, rhythirucality is still felt to be a basic principle of the music and its performance. The gap between strict metronomic rhythm and loose 'psychological' rhythm also exists in language, where there are even more factors to interfere with the ideal of isochronism. For example, the duration of the measure (corresponding to the musical bar) tends to be squashed or stretched according to the number of unstressed syllables that are inserted between one stress and the next, and according to the complexity of those syllables. In this, a speaker of English is rather like a would-be virtuoso who slows down when he comes to difficult, fast-moving passages of semi-quavers, and accelerates on reaching easy successions of crochets and minims. Although some people reject the principle of isochronism because of the lack of objective support for it, I shall treat it here as a reasonable postulate without which a meaningful analysis of rhythm cannot be made. What we call' stress', by the way, cannot be merely reduced to the single physical factor of loudness: pitch and length also have a part to play. Stress is an abstract, linguistic concept, not a purely acoustic one.
7.2.1 The Measure: the Unit of Rhythm
As the rhythm of English is based on a roughly equal lapse of time between one stressed syllable and another, it is convenient, taking the comparison with music further, to think of an utterance as divided into 'bars' or (as I have already called them) measures, each of which begins with a stressed syllabic, corresponding to the musical downbeat. A number of unstressed syllables, varying from nil to about four, can occur between one stressed syllable and the next, and the duration of any individual syllable depends largely upon the number of other syllables in the same measure. If we assign the value of a crochet to each measure, then a measure of three syllables can be approximately represented by a triplet of quavers, a measure of four syllables by four semi-quavers, etc. This method of rhythmic analysis,5 which is not to be considered a method of'scansion' as usually understood, is illustrated in these two passages of rhythmically free poetry:
World's strand,1 sway of the | sea,
Was this article helpful?