Division of Utterances into Intonation-Groups

Analysis of English utterances into intonation-groups shows that they are co-extensive with a stretch of speech of various grammatical nature: an independent sentence, a principal or a subordinate clause, two or even more clauses, a group of words or even one word. Co-extensiveness with a sentence is typical of only a small portion of speech material (about 17 %, according to experimental data). An intonation-group corresponding to a grammatical sentence is marked by specific characteristics of tone, stress and duration, serving to express semantic completeness and independence — the relevant features of an utterance. An intonation-group of this kind is defined as a simple tune.

Most grammatical sentences are prosodically expressed by a combination of intonation-groups. These combinations have a specific function of a double nature: on the one hand, they present information in the form of relatively separate semantic items, and on the other hand, they make up a communicative whole (entity) out of these separate parts. Utterances which are composed of more than one intonation-group form a combined tune.

Some sentences lend themselves to be subdivided more readily than others. Long sentences, most naturally, break up into smaller parts in spoken language. Their division is based both on physiological convenience (an intonation-group is normally a breath-group) and on the complexity of information being conveyed, e.g.:

After a long boring wait | I eventually boarded my plane.

Of the two factors - physiological convenience and complexity of information - semantic reasons are overriding in importance. Through intonation division the speaker can make several items stand out as more or less independent parcels of information in a short utterance, too, increasing thereby the general prominence of the utterance,

Nobody | could deny it.

Another major characteristic involved is the syntactic structure of an utterance. The number of intonation-groups in utterances of the same length may often vary precisely because of the peculiarities of their syntactic structure, which may either presuppose prosodic division as an obligatory feature or, vice versa, 'forbid' it, or else (as a third and most frequently occurring variant) allow of two options: with or without an intonation boundary between the constituents of a sentence.

Prosodic division is typically optional in expanded simple sentences with adverbial modifiers of different kinds, complex sentences with object, relative or attributive clauses and some others. The grouping of words within a message into longer or shorter sections and the placement of an intonation boundary in such cases is largely a matter of the speaker's semantic interpretation of an utterance, as well as his communicative intention. As a result the same written sentence read aloud by different people may have a different number of intonation-groups. E.g.:

Many working mothers do not have time to cook. Many working mothers | do not have time to cook.

She has learned to keep quiet about her personal relationships. She has learned to keep quiet | about her personal relationships.

Often the number of intonation-groups is the same, but the location of their boundaries varies. E.g.:

Los Angeles | is well known | for both the high level of its air pollution | and the efforts made to control it.

Los Angeles is well known | for both the high level of its air pollution | and the efforts | made to control it.

An intonation boundary is obligatory, or, at any rate, highly probable in complex sentences with subordinate clauses of condition, cause, time (in pre-position to the principal clause), concession, result, comparison (particularly, when there is an adverbial modifier of manner in the principal clause) and some others. E.g.:

Since you refuse to help, | I must do it alone.

In spite of the rain and bitter cold | they all came in time.

Strictly speaking, there is no rule forbidding a pause in any place within an utterance (cf. the so-called hesitation pauses), but from the point of view of syntactical predicta­bility certain positions in an utterance display a very small probability of a break. Thus, e.g. the subject of a sentence expressed by a personal pronoun is but seldom separated from the predicate; a preposed attribute is usually closely linked to the noun, etc.

The choice of a number of intonation-groups in an utterance also depends on the type and form of speech. In a dictation, for instance, an utterance is divided up into smaller sections than in any other kind of reading, and spontaneous speech is characterized by uneven length of intonation-groups, and their boundaries are less predictable from the syntactic structure than in reading aloud.