Aarhus University Seal / Aarhus Universitets segl

Prosody

Prosody covers everything that has to do with sounds in language, except for the individual sound segments. This is. e.g., stress, intonation, volume, length, speed and boundary signals. We also put voice quality, personal voice features and specific uses of the voice under the category of prosody, but not the Danish stød (glottal constriction) or word tones (for more, go to Our method).

The anatomy and the physics behind prosody

Many of the sounds we make when we speak are produced by air flowing from the lungs up through our vocal chords, which can then vibrate. The speed with which the vocal chords vibrate determines the pitch (the intonation or frequency which is measured in Hz). It is the width of the fluctuation of the vocal chords, the amplitude, which determines volume. The volume can be lowered or boosted, depending on other factors as well. By tightening (parts of), releasing, or opening the vocal chords or other parts of the vocal organs, we change the voice quality, which entails that we can describe the voice as ‘smiling’, ‘laughing’, ‘tense’, ‘pathetic’, ‘airy’ and so on. We can also change the sound expression by moving the speech organs faster or slower, i.e. speaking faster or slower and, of course, by not speaking.

By relating the physical factors to the production of speech sounds, we can begin to understand what stress, intonation, speed, prosodic boundaries, and other prosodic signals are:

Stress

Stress has to do with syllables. Some syllables are stressed, others are not. Stress is a rather complex concept. In Danish, stressed syllables last longer, have more pitch changes and more volume than unstressed syllables. In our transcriptions, we distinguish between no stress and more or less stress. This is indicated by underlining more or less of a syllable (se Transcription conventions).

One can distinguish between lexical stress and “utterance stress”. Lexical stress is the stress in words with more syllables when they are pronounced individually, and utterance stress refers to the syllables which are given more stress in coherent speech. The latter are the ones which have our interest because our research is based on coherent speech. When one looks at utterance stress, the speech can be separated in stress groups (feet). A stress group lasts from one stressed syllable to right before the next one.

Utterance stress is central for understanding the meaning of what is said. There is, e.g., a difference between saying “'stå op” (stress on the first word) and “stå 'op” (stress on the last word) in Danish, the former meaning ‘be in a standing position’ and the latter ‘get up’. The utterance stress also has important morphological consequences. For the varieties of Danish that have stød, there is both a version with and without stød that is used depending on whether a word has stress or not. This is specific to verbs as a word class (for more, go to Verbs), but also words from other word classes change in stød and other pronunciation features depending on stress.

Intonation

Intonation is the melody, or pitch, of an utterance.
In our transcriptions we show intonation by applying arrows at the end of a given utterance. This only depicts the pitch movement at the end of an utterance (from the last stressed syllable till the end of the utterance). Sometimes we also show sudden jumps in intonation with arrows which point up or down. This of course does not do justice to the intonation patterns. Therefore we have in some articles (for more, go to own production and writings on Danish talk in interaction) made pictures of the intonation in the program Praat. In this way, we can show how the tone moves through the utterance.

Even though some research har been carried out on intonation in Danish talk-in-interaction, there are still many things we do not know. It is, e.g., still unclear Danish has a specific questioning. On the other hand, it is evident that the functions of ‘little words’ which can stand alone (Interjections and particles) very much depend on their intonation.

There is therefore no doubt that intonation has its own place in a description of the grammar of talk-in-interaction.

Speed

We can hear when someone speaks fast or slowly. Therefore we also indicate in our transcriptions when someone slows down or speeds up their talk. If someone pronounces many syllables per time unit, one can choose to have fewer stressed syllables which gives a different impression (and perhaps also a different grammatical function) than if the same amount of syllables are produced per time unit but with more stressed syllables.

It is not simple to measure the speed of talk. This is because it is unclear whether it is the amount of individual sounds, syllables, words or stressed syllables per time unit which decides whether we hear talk as fast or slow. In Danish we also reduce sounds and syllables, which entails that more words are produced per time unit, but not necessarily more sounds or syllables.

Boundaries

Prosody contributes to our perception of beginnings and endings of utterances, that is, it makes boundaries for the speech. There are many suggestions as to which prosodic features function as boundary signals, but also this aspect is so complex that we do not yet have a clear picture of how prosodic boundaries are being signaled. Here is the short version:
You can often clearly hear when a new utterance begins. When you hear this, we talk about pitch reset.

The most important feature about pitch reset is that the intonation in one intonation unit is higher than in the preceding intonation unit. The reason for this is that intonation generally falls through the course of an utterance in Danish. Furthermore, a faster pronunciation of non-stressed syllables prior to the first stressed syllables can cause one to hear something as a pitch reset.

In some transcriptions, pitch reset is depicted with a capital first letter. In Samtalebankens transcription conventions, a specific symbol; a half circle with an arrow pointing up, is made for exactly this. We have chosen not to depict pitch reset in our transcriptions, mostly because we don’t want to make them complicated to read. We can use Samtalebankens symbol though, if pitch reset becomes specifically important in a given context.

When it comes to endings of utterances, a clearly falling intonation to the high end of a speaker’s register – and sometimes also a clearly rising intonation to the high end of a speaker’s register – may be heard as a signal of possible completion. This pattern does not seem to be used much in Danish though, and for the most part, utterances reach completions without the intonation being a clear marker of this. At the end of an utterance, the speed of talk is often slowed down and so is the volume. Also features such as creaky voice and hearable exhaling often occur at utterance endings. The speaker can of course also pause when an utterance reaches a completion, but pauses also appear in the middle of utterances. A next speaker does not, generally, wait to take the turn until he/she is sure that the first speaker is pausing, and, therefore, we cannot use pauses as a clear indication of a boundary.

Prosodic signals about specific functions

Prosody can signal the function of an utterance. Speakers can, e.g., imitate other people’s voices when quoting them, with a tone of voice that both indicates that they are quoting and what type of person or mood they are imitating. For this, all the prosodic resources (intonation, volume, voice quality, speed) can be used, in ways which have only been cursorily described.

When we speak, we invariably send signals that concern our states of mind and what we wish from our conversation partners. We can sound happy, sad, excited, indifferent, intimate, inviting, flirtatious, cold, angry and so on. We can name these signals affective or state of mind-signals. With these, we reveal who we are and what relationship we wish to have with co-participants. Affective signals also have a situational function. They can, e.g., contribute designing an utterance as serious, humorous, ironic, sarcastic, intimate, etc.

In the future, we will make entries about significance of prosody for other particles, for the understanding of utterances (does Danish talk-in-interaction have a “questioning intonation”?), the significance of stress for understanding talk as coherent and for distinguishing different grammatical constructions, pitch reset, prosodic boundary signals, etc. 


Entries under Prosody


Further reading

dialekt.dk is a scientific website about Danish dialects. On this website you can listen to different examples and it gives you an impression about differences in prosody (so far only in Danish).

 

Kirk (2008) is a thorough guide to Danish pronunciation meant for teachers who teach adults who are learning Danish  as a second language (Danish).

 

Steensig (2001)contains a thorough section about the role of prosody in relationship to turn construction in Danish conversation language. This is an academic text (Danish).

 

Grønnum & Tøndering (2007) is an examination of Danish questioning intonation. This is an academic text in English.