Mirth measurement

Mirth Measurement: A New Technique

Mirth Measurement: A New Technique

Michael Mair and John Kirkland

University of London Institute ot Education

The approach to the study of mirth described here is part of an investigation both of movement and of sound patterns generated by participants in interaction. Let us first explain about some of the data collected including a description of how these were taken, and then go on to explore their significance in a broader context.

METHOD

At present our equipment can deliver four types of measurement. Briefly these include: basal intonation contour; voice onset and offset; total speech pressure wave; and, finally, manual frame-by-frame plotting of face movements.

The first three are acoustic variables. These are delivered by use of the laryngograph tape-recorder, and ultra-violet pen-corder. The laryngograph monitors each opening and closure of the vocal folds by measuring impedance change between disc-electrodes placed on each side of the thyroid cartilage. This signal is electrically converted to a graphic plot showing fundamental frequency (Hz). Paper speed is 10 cm per second. Thus the acoustic equipment delivers speech pressure wave, the original signal, and the conversion. These traces are displayed simultaneously (see Fourcin, 1974; Fourcin & Abbertan, 1971).

Figure 1 shows fundamental frequency traces from two mirthful episodes. The visual record is obtained by monitoring a single point an the face (medial canthus of an eye or tip of nose) as it changes its position between successive single frames on a video-corder. The equipment used is a Sanyo (Model 1100 SL) which delivers stable single frames. These frames are numbered by including a digital clock during recording. The time interval between frames is chosen as 0.1 second, because inertia of the head makes notable changes of direction impossible within a shorter time period. Additionally, critical fusion frequency is around this speed and makes perception of faster changes unlikely. For a single frame the selected face-point can be established by noting co-ordinates (x,y) in two-dimensional space. By following this point across successive frames we include time (t). These data are represented as x/t; y/t; and x/y.

To eliminate parallax errors when obtaining face-point co-ordinates a transparent grid is first applied to the front of the video monitor. The film is then viewed through a pinhole mounted at a fixed but convenient distance from the monitor (in our case 1m). This pin-hole is a retina immobiliser. One alternative would be to incorporate grid and cross-wires into the input video signal but this is expensive.

Subjects were chosen who already knew one another. Our discussion is based on data taken from three subjects. Two mirrors arranged to form a convex-angled V were placed on a table between seated interactants who could still maintain eye contact. This device enabled us to film simultaneously two subjects with a single video-camera. Two laryngograph electrodes, each the size of a two-pence piece, were attached. An ordinary microphone was used as well. It enabled later location of critical tape sections. Thus each subjects acoustic signals were fed into twin-tracks of a single tape-recorder. Selected portions of audio-tapes were transcribed by the ultra-violet pen-corder.

Subjects were kept naive about the purposes of the study. Their sole instructions were that they should 'chat'.

DATA

We are considering mirth from the aspects of fundamental frequency, rhythmicity, supra-glottal modification, movement, and synchrony. For completeness, we have also included other physiologic eruptions, speech, and a miscellaneous category comprising such events as 'hums', 'moans', etcetera. The categories are demarcated for convenience although in real communication they merge. However, if we consider them exhaustively from each aspect we may see whether they survive in any sense. We shall conclude 'mirth' in this way.

Example 1

The acoustic variables are shown in Figure 1, and the head movement in Figure 2. These data were obtained from an exchange between interactants, demarcated from the surrounding conversation by silence, and separate from it in topic as well. This exchange consists of a question: Male voice: 'Great fun breaking up polystyrene glasses isn't it?' and an answer: Female voice: 'Yeah, make a very great mass can't you?' The male laugh occurred over the word 'glasses', and the female laugh preceded her utterance and occurred synchronously with the male laugh. The Figures show:

1. Fundamental frequency (C): the male laugh occurred over the ward 'glasses'. In all, his utterance consisted of two 'S' shaped falls, linked by a slow rise. The laugh disturbed, but did not destroy, the shape of the slow rise, which continued as a trajectory through it. The female laugh also has pitch, which was high, as was the commencement of her utterance.

2. Rhythmicity: the male laugh turned the word 'glasses' into a stutter. The female laugh also consisted of more than one pulse.

3. Supra-glottal modification: the male laugh disturbed the articulation of the word 'glasses'. The female laugh was not accompanied by an attempt at articulation.

4. Non-articular movement: our technique delivers a record of gross head movement only. In Figure 2 we see that the male went into a relatively immobile head position from frame 9 to 17. His escape from immobility occurred at about frame, 18, which was close to the onset of the laugh, and he returned to relative immobility at frame 36. The female entered the episode in a condition of relative immobility. Her movement also coincided with her laugh, and returned to relative immobility.

5. Synchrany: this refers to the extent to which they moved together during the laugh. They initiated the laugh movement within 0.4 second of each other. The laugh vocalisation was more closely synchronised.

Example 2

This is a record of a solo female whose laugh occurred in an interaction involving three people (not shown). It was sparked off when the experimenter attempted to join the conversation. It consisted of a laugh, followed by the comment 'Hm, bring you into the conversation', followed by an unclassifiable vocalization. Figures 1 and 2 show:

1. Fundamental frequency (C): the pulses followed a falling contour which was continued into the beginning of the utterance.

2. Rhythmicity: the laugh had a three pulse, two pulse, three pulse sequence.

3. Supra-glottal modification: the laugh was free from supra-glottal modification.

4. Non-articular movement: there was an abrupt onset of movement at the vocalized part of the laugh, and the subsequent movement is represented in Figure 2 by large loops which become a more detailed shape when she is speaking (frame 28). Rapid movement is resumed over the final syllable, when she also looks down.

5. Synchrony: not applicable.

DISCUSSION

Our examples are evidently too few to prove any interpretation. However, we do not intend to deny ourselves this opportunity to put forward a theory, albeit speculative, which links some features of these examples, and places them in the wider context of some observations on speech. Let us list the features of mirth from our examples.

1. Fundamental frequency: this is related to the contour of the utterance in which the mirth is embedded.

2. Rhythmicity: mirth's rhythms are simple and pulse-like.

3. Supra~glottal modification: articulation is either disturbed (the stutter), or absent.

4. Non-articular movement: mirth involves gross body movement with abrupt onset.

5. Synchrony: the onset of the movement was near synchronous for the interactional example.

How do these features compare with the same categories manifesting during speech?

1. Fundamental frequency: the relationship between fundamental frequency and perceived intonation is complex, but close. Certainly, the auditory affect of the laryngograph output is very similar to the perceived intonation of speech. Fundamental frequency as measured by the laryngograph is a monitoring of motor activity in the muscles regulating the vocal folds. The question why mirth should participate in the frequency shape of the utterance cannot be approached separate of the question why speech should have a fundamental frequency shape at all. Observation of the fundamental frequency contour for numerous utterances demonstrates smooth shapes, which continue across pauses and silences to be continued where they would have got to if the pauses had not occurred. We propose the term 'trajectory' to express the slower wholeness of intonation tune over utterances (there is also a finer embroidery corresponding to segmentation). It is the shapes themselves that suggest they be considered as trajectories.

2. Rhythmicity: the complex rhythms of speech have attracted much study. At this point, it is sufficient to note the contrast between them and the simple rhythmic vocalization of mirth.

3. Supra-glottal modification: this is the articulatory component of speech. Again, at this paint it is sufficient to note that it is very detailed, and its study has an enormous literature.

4. Non-articular movement: this also is very detailed, and there are many approaches to its study (e.g., Ekman, 1969). We have noted that it is going with speech, and this realization follows the work of Condon and Ogston (1967) with Kendon (1972). It is in contrast to the gross movement of mirth.

5. Synchrony: the question of synchronous movement between participants in interaction is also being explored by us; we have many examples of it. Again, we follow the work of Condon and Ogstan (1967).

It is the fate of students of human communication to have to demarcate aspects of it for study, while being aware that in the process of communicating, all these aspects interlock. Perhaps a conference on humour can be the place to present a theory which might arouse mirth in another context?

We suggest that when studying communication - when trying'to make sense of it - we are in fact studying the process by which sense is made; that when people are communicating, they are trying to make sense to each other. Trying to make sense of the process itself differs from the sorts of sense that people habitually try and make only in subject, and setting. We propose to call the unit of sense the 'model'. There are numerous other concepts with which this can be aligned, for example, the 'word group' (O'Connor & Arnold, 1974), the 'sentence' (Goldman Eisler, 1972, after Wundt), the 'tone group' (Kendon, 1972), the 'act' (Thom, 1975). Each seeks to delineate a unit of speech which is both an information unit, and a physically demarcatable stretch of speech.

Part of the problem is that for such a model (a realization of a state of affairs) to be communicated, nothing need actually be said. A wink can do it, or a chuckle. Even a silence can 'be so loud so as to drown out the scuffle of feat' (Birdwhistell, 1971). Such a silence clearly communicated a state of affairs or, as we prefer to call it, a model. That such as these can 'click' a model into existence points to two features of models. They are to a greater or lesser extent implicit, and they are shared - people are immersed in them, rather than exchanging them like counters. Of course, in formal situations (like this), the passage seems more one-way. But even here we share much more than my individual utterance contributes. We call these models 'interactional', to stress that in colloquial conversation, the model is a joint production, and evolves by the collaboration of the participants.

The study of colloquial conversation performed by this technique might give verification to this view. Again, we make our list:

1. Intonation contour: this is far from random. O'Connor and Arnold (1974) suggest that there are but ten common tunes in colloquial English. The nucleus of the utterance coincides with accent and pitch change, and is always semantically important. Thom (1975) has a predatory theory of language, in which it is seen as virtual action, and grammar as being the same as the syntax of action. His archetypal morphologies (verb types, such as capture, emit, almost) look uncommonly like some common intonation contours. In colloquial chat, we complete each other's utterances, continue each other's contours. Together we make the model. It is as if the intonation contour is always just appropriate to where the model has got to in its elaboration.

2. Rhythmicity: if we consider contour as trajectory in this way, then we might suggest that it is the shape of the trajectory which divides up time into rhythm. Rhythm is not perfectly metrical, yet it is predictable. This view might contribute to the thoughts on rhythm offered by authors such as Abercrombie (1969) and Allan (1973).

3. Supra-glottal modification: Candon and Ogston (1967) have claimed that micro-movements go with speech to a detail as fine as the phoneme. Liaberman (1972) notes that for phonemes to be perceived sequentially would be beyond the discriminatory capacity of the ear. Ivimay (in press) points out the dramatic short cuts taken in articulation during spontaneous speech, and our studies confirm this (for example, 'sort of' becomes Istvl, or less). Micro-movements do certainly occur to a detail equal to the actual change points in articulation and Kendon (1972) has suggested that it is as if speech and the movements that immediately accompany it are under the same controlling mechanism. But intonation change can occur with this detail as well. So:

4. Non-articular movement, and 5. Synchronous movement cannot be approached theoretically separate from all the other categories above. Furthermore the 'controlling mechanism' which effects synchronous movement might best be considered supra-individual.

The comparison of speech with mirth shows up what mirth lacks. Complex rhythms are replaced by simple spasmodic rhythm, articulation (supra-glattal modification) is disturbed or abandoned. Detailed body movement becomes gross body movement. Frequency contour is at first maintained but even this is lost in extreme cases. The interactants are in a state of uncontrol (Douglas, 1975), on the floor, helpless, pouring secretions from eyes and mucous membranes (vague nerve), sometimes incontinent. Only the synchrony of onset is preserved. It is catastrophic, and the way that the punchline of a joke can do this to the rest of the model preceding it is reminiscent of the way a Necker cube abruptly flips from one interpretation to another (as Gestalt theorists suggest). Our interactants must recover. Why we should seek out this experience is a weighty question indeed, but it seems that mirth may survive as a demonstrable pattern of behaviour.

FOOTNOTES

Support by a grant from the Social Science Research Council is acknowledged.

Now at Education Department, Massey University, Palmerston North, New Zealand.

The laryngograph was devised in the Phonetics Department, University College, London. It is available from Laryngograph Ltd., 24 Highclare Drive, Hemel Hempstead, Herts., HP3 8BY.

REFERENCES

See Bibliography for publications on humour, laughter and comedy

Abercrombie, D., Voice qualities. In: N.N. Markel (Ed.), Psycholinguistics. Dorsey Press, Homewood, Illinois (1969).

Allen, C.D., Segmental timing and control in speech production. Journal of Phonetics, 1,-219-237 (1973).

Birdwhistall, R.L., Essays on Body-Motion Communication. Penguin, Harmondsworth (1971).

Candon, W.S. & Ogston, W.D., A segmentation of behavior. Journal of Psychiatric Research, 5, 221-235 (1967).

Ekman, P., The repertoire of non-verbal behaviour: categories, origins, usage, and coding. Semiotice, 1, 449-498 (1969).

Fourcin, A.J., Laryngographic examination of vocal-fold vibration. In: 0. Wyke (Ed.), Ventilatory and Phonatory Control Systems. Oxford University Press, London (1974).

Fourcin, A.J. & Abberton, E., First applications of a new laryngograph. Medical and Biological Illustration, 2, 172-182 (1971).

Goldman Eisler, F., Pauses, clauses and sentences. Language and Speech, 15 (1972).

Ivimey, G.P., The perception of speech: An information processing approach. Teacher of the Deaf, in press.

Kendon, A., Some relationships between body motion and speech. In: A.W. Siogman & B. Pope Eds.), Studies in Dyadic Communication. Proceedings of a Research Conference an the Interview. Pergamon, New York (1972).

Liaberman, A.M., Perception of the speech code. In: E. David & P. Dense (Eds.), Human Communication, a Unified View. McGraw-Hill, New York (1972).

O'Connor, J.D. & Arnold, G.F., Intonation of Colloquial English. Longman, London (1974).

Them, R., Structural Stability and Morphogenesis. (Translated D.H. Fowler). Benjamin, Massachusetts (1975). EMBED Unknown

EMBED Unknown

_966074723.unknown

_966076588.unknown