Jürgen Trouvain, Khiet P. Truong Published online on: 28 Feb 2017

Transcription

1 This article was downloaded by: On: 10 Jan 2019 Access details: subscription number Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: Registered office: 5 Howick Place, London SW1P 1WG, UK The Routledge Handbook of Language and Humor Salvatore Attardo Laughter Publication details Jürgen Trouvain, Khiet P. Truong Published online on: 28 Feb 2017 How to cite :- Jürgen Trouvain, Khiet P. Truong. 28 Feb 2017, Laughter from: The Routledge Handbook of Language and Humor Routledge Accessed on: 10 Jan PLEASE SCROLL DOWN FOR DOCUMENT Full terms and conditions of use: This Document PDF may be used for research, teaching and private study purposes. Any substantial or systematic reproductions, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The publisher shall not be liable for an loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

2 1 Introduction Studies of laughter are spread over multiple scientific disciplines of which many have nothing or nearly nothing in common: from conversational analysis and interactional linguistics to emotional psychology, from phonetics and speech technology to physiology, from evolutionary biology to cultural studies. There is not a single established laughter research community but several interest groups that were and still are emerging in parallel. Thus, the many different facets of laughter lead to a big diversity of research foci of which only a selection can be treated in this chapter. 1.1 Humor and Other Sources of Laughter Laughter and humor are frequently used synonymously. There is no doubt that humorous action can lead to laughter (and smiling); however, not each jocular remark is marked by laughter by either the recipient or the speaker. The reasons why laughter and smiling are evoked are manifold and most of them have nothing to do with humor. Apart from amusement and joy, laughter can be a display of nervousness, positive surprise, hilarity, pleasure, non-seriousness, and affiliation, but also maliciousness or as a face-saving or threatening action (e.g., O Donnell-Trujillo & Adams, 1983). Laughter in the aforementioned circumstances usually occurs in a spontaneous and unplanned way. But there are also forms of non-spontaneous laughter for instance, faked laughter and acted laughter (e.g., on stage). Laughing in music (e.g., in opera songs) represents a particular form of staged laughter. Further special cases include laughter with extrinsic stimulation, e.g., by tickling or laughing gas (formerly used in anesthesia). Also, vocalizations in laugh yoga groups can be considered as extrinsically stimulated. In addition, there are forms of pathological laughter such as risus sardonicus. 1.2 Laughing and Smiling Another often used synonym pair is laughing and smiling. A formulation like she is laughing with me sometimes actually means she is smiling with me. Laughing is usually expressed Laughter Jürgen Trouvain and Khiet P. Truong

3 Laughter acoustically and visually, whereas smiling does not necessarily make use of the acoustic channel. The facial expression of most forms of laughing is more complex than that for smiling. Nevertheless, smiling and laughing share some functions, e.g., as an expression of joy. Forms of smiling are often divided into felt (or so-called Duchenne) smiles, e.g., when someone is delighted, and non-felt smiles, e.g., to signal politeness (Ekman & Friesen, 1982), though there seems to be different interpretations of smiling across cultures (Thiebault et al., 2012; Krys et al., 2016). There are clearly parallels between smiling and laughing, and smiling is sometimes considered the little brother of laughing: compare the respective words in German (lachen and lächeln) and French (rire and sourire). Since Darwin (1872) it is a matter of debate whether smiling and laughing are two distinct categories or two ends of the same scale. 1.3 Laughter Always Haha-Like? What does the prototypical laugh sound like? Many native speakers of English and German would probably answer haha, Italians maybe with ahah, and native speakers of Spanish might say jaja. For these tamed and stylized forms of sound imitation the phonological constraints of each language were applied. Here, words or word-like units were conventionalized to generally denote laughter an acoustic communicative event that is not a word. Sometimes laughter is considered to belong to the word class of interjections, i.e., particles with an affective component, often with a sound-imitative character. However, laughter as such, i.e., in its raw and untamed form, should be considered a non-verbal vocalization along with breathing sounds, sighing, or clicking (Trouvain, 2014). Although laughter is a salient acoustic behavior in everyday communicative situation, there are comparably few studies on its acoustics, perception, and physiology. As a starting point of scientific descriptions Darwin (1872) can be considered. 1.4 Laughter and Emotion Laughter is often seen as an expression of emotion and linked to positive feelings. Indeed, prototypical vocalizations to express happiness generally use forms of laughter (Belin et al., 2008; Sauter et al., 2010). However, things are a bit more complicated. The set of (basic) emotions varies between studies (including joy, happiness, pleasure, elation and more). Moreover, there seem to be various forms of happiness (Sauter, 2010). Looking at emotions in dimensions like intensity (strong mild), valence (positive negative) and power (dominating subordinated), laughter can appear at quite different and sometimes contradicting coordinates. Laughter expressed to underline a malicious thought is rather negative, as it is with a laugh uttered in a very sad setting (e.g., somebody is talking of the deceased partner). Another problem with the term emotion is that it is often used as an umbrella term for affective states, i.e., for a short-term inner state but also for a medium-term inner state like a mood, and also as reflections of attitudes towards things and people who sometimes express a stance. Moreover, in real situations emotions such as pure joy are comparably rare. These short-term affective states are infrequently full-blown emotions, and frequently mixed with further affective components. 1.5 The Social Nature of Laughter Laughter seems to mainly fulfill a social function. Laughter in particular, mutual laughter, creates a form of social bonding; that is, it is a display of affiliation, and can usually be associated with 341

4 Jürgen Trouvain and Khiet P. Truong a pleasurable atmosphere (Jefferson et al., 1987; Grammer & Eibl-Eibesfeldt, 1990; Smoski & Bachorowski, 2003; Mehu & Dunbar, 2008). Through observational studies, Mehu (2011) found that high intensity laughs mostly occurred between individuals of the same age, supporting the view that laughter could be involved in bonding and cooperative signaling. Adelsward (1989) found more mutual laughter in successful job interviews than non-successful ones. In the same study, more total laughter was found in conflict dyads but more mutual laughter was found in cooperative dyads. The occurrence of mutual laughter was also investigated in dyadic interactions between friends and strangers by Smoski and Bachorowski (2003). They found that mutual laugher occurs more in dyads composed of friends than dyads composed of strangers. This result is in line with their view that mutual laughter may support or reinforce mutually pleasurable affective experiences, which is probably more often the case with strangers. If laughing together can be seen as a sign of alignment and affiliation, then not laughing with someone could indicate disalignment or disaffiliation. As pointed out by Sacks (1992), laughter can activate a normative framework in which laughter should be shared. In certain situations, not sharing the laughter could be seen by the participants as a problematic action. Similarly, Schenkein (1972) suggests that by withholding a laugh from a slot in which one is specifically appropriate, as in the slot following a speaker s joke, a person can disaffiliate and cause trouble for this speaker. Others have pointed out the role of mutual laughter in managing the flow of the conversation. For instance Holt (2010) investigated patterns in the distribution of shared laughter in dyadic interactions. It was found that shared laughter often precedes the introduction of a new topic, thus serving as a topic termination cue. The sequencing of shared laughter has been another focus of attention. Jefferson et al. (1987) noted that laughter sequences can be seen as forms of adjacency pairs, similar to greeting return greeting pairs. A first laughter instance can serve as an invitation that can be accepted by the recipient by placing a subsequent laughter instance (Jefferson, 1979). In addition to dyads, Glenn (1989/2003) studied the occurrence and sequencing of shared laughter in naturally occurring multiparty interaction where there are more than two speakers present. He was more focused on the question of who laughs first. He found that, in general, in dyadic interactions, the current speaker laughs first. In general, in multiparty interactions, it is someone other than the current speaker who laughs first. Parallels can be drawn to a stand-up comedian situation where the audience usually laughs first, followed by the comedian. The strong bonding effect obviously leads to laughter having a strong bias to be contagious. Probably everybody has experienced for her/himself an instance of being infected by laughter of others. This fact has led to an attempt at exploitation by the playback of prerecorded laugh samples of an audience in TV series at certain locations in the discourse. 1.6 Development of Laughter Is laughter unique only to the human species? There seem to be various species enabled to show a behavior that is similar to human laughter, among them dogs, hyenas, parrots, chimps, bonobos, and orangutans (Davila-Ross et al., 2009). For some species, laughter-like vocalizations are mere imitations of the acoustics of laughter just heard. For other species, laughter triggers social interaction. In a study by Davila-Ross et al. (2014), chimpanzees responded to imitation by a humanoid robot and requested responses from it. The phylogenetic origin of smiling is often referred to the silent bared-teeth display, whereas the relaxed open-mouth display is also known as play face. In a study with chimpanzees Walter 342

5 Laughter and Dunbar (2005) observed that dyadic play bouts are significantly longer when the openmouth display is bidirectional. Also the degree of affinitive behavior increased after both displays. Their data suggest social bonding as a similar evolutionary function for both displays, which could explain convergence of the two displays in humans. From an ontogenetic perspective it can be stated that there are huge differences in the laughing behavior between individuals. This concerns, among other factors, the number of how many times people laugh, in which situations they laugh and the phonetic makeup of their laughs. The idiosyncratic character of laughter is also used when displaying specific characters in movies, dubbed voices or in animated characters. It can also be assumed that the individual laughing behaviors change over a lifetime, and that laughter, as other (vocalic) behavior, is subject to social influence. Although universal, laughter is also culturally affected in the sense that people with hearing abilities influence each other in the way of laughing. A study by Makagon et al. (2008) showed that the laughs of congenitally deaf subjects had a smaller range of phonetic variation than those of subjects with normal hearing. 1.7 Laughter in Spoken Interaction Although laughter can be considered and investigated as an autonomous event, it is the context in which the laughter occurs that allows its interpretation. Usually laughter is contextualized within spoken interaction. This entails, for instance, the exact timing of laughter of the interlocutors, the phonetic quality and duration of laughing as well as the embedding of laughter in speaking situations and cultures. Special issues and article collections on laughter in interaction (Wagner & Vöge, 2010; Glenn & Holt, 2013) and specialized workshops on phonetic topics of laughing (Trouvain & Campbell, 2007) provide details of this complex intertwined relationship of articulated language and the vocal but non-linguistic signal of laughter. 2 Core Issues and Topics This section focuses on the most important topics of laughter research, starting with methods of how data can be collected and how laughter could be observed. What follows is an overview of possible ways of describing and transcribing laughter. Measurement methods of the shape and sequencing of laughter end this section. 2.1 Collecting and Observing Laughter Laughter is a non-verbal phonetic activity that usually occurs in conversational interaction with an interlocutor. In contrast to this we can state that many studies on the acoustics of laughter were not based on conversational settings but settings in which actors produce preselected laughter categories (Habermann, 1955; Szameitat et al., 2009) or in which subjects watch funny video clips, either alone (Urbain et al., 2010) or with another person (Bachorowski et al., 2001). Everyday situations would be the ideal setting for observing laughter. However, this ideal setting will be suitable for scientific purposes only in a few exceptional cases. The approach of just listening to situations in which laughter occurs with subsequent taking of written notes (e.g., Provine, 2000) is at least very doubtful and cannot be recommended. The observer is either not fully informed about the context or is part of the conversation, which provokes a strong bias. Because acoustic data are difficult to record in such a situation, most details are not observable at all and do not allow any reproducibility. As soon as data can be recorded we face the observer s paradox (Labov, 1972); i.e., it cannot be excluded that people behave differently when they feel 343

6 Jürgen Trouvain and Khiet P. Truong observed. Covert recordings would overcome this problem but cannot be considered for ethical reasons (recordings in a forensic context may represent an exception of this rule). Another often-encountered problem with existing recorded data is that interlocutors are often audible on each other s recorded audio channel. Because interlocutors of a conversation often overlap with their speech and particularly with their laughing, this could pose problematic issues for acoustic analyses. Important information can be masked, for instance, when the beginning and the end of a laugh of a specific person is not audible or the laugh is not audible at all. An optimal observation setting would involve a separation of the acoustic channels of the different speakers. However, most recordings used in conversational analysis studies were carried out with just one microphone, where the aforementioned acoustic details possibly remain unknown. Technically, optimal recordings are definitively a plus when analyzing laughter. Thus, recordings taken from media (radio, television, shared videos on the web) could serve as a rich source of data. But there are two main problems here: the researcher has a reduced or complete lack of control of the situational context and technical manipulations cannot be excluded, or the conversational atmosphere was highly artificial compared to the conversation before the recording. The next problem is that speakers in media behave in a certain conversational manner that makes them professional. They act in a given role similar to actors performing and the product has the effect of leaving a more or less authentic impression. There are various forms of eliciting laughter in experimental settings. For instance, Grammer and Eibl-Eibesfeldt (1990) audio- and videotaped male female encounters when the subjects felt unobserved. Nwokah et al. (1999) recorded mother child interactions in the lab. Truong and Trouvain (2012a) analyzed various corpora with conversational speech, which were not specifically recorded for laughter analysis. These corpora were either task-based, e.g., interlocutors played a spot-the-difference game, or they contained speech of real meetings of a research lab where each interlocutor wore a microphone. To summarize, the systematic observation of laughter is by far a non-trivial task. What becomes clear after studies with various styles and in various situations is that laughter is not a strict, staccato-like voiced voiceless alternation alone but more a bundle of very different phonetic and diverse complex forms (cf. Trouvain, 2003; Chafe, 2007). 2.2 Describing and Transcribing Laughter How many different types of laughter can we find? Such a question of course depends on the typology you are searching for. When looking for lexical variants for laugh in the English language we could come up with the following list: 1 giggle, cackle, guffaw, chuckle, chortle, snicker, choke, titter, bray, howl, horselaugh, mock, groan, bellow, hoot, bleat and neigh. Other languages will definitively show other lists of synonyms and related words. This semantic field of laugh can be extended by emotional and attitudinal labels for laughter such as joyful, cynical, heart, mirthful, sad, dirty or belly. A more phonetic impressionistic approach goes back to Grammer and Eibl-Eibesfeldt (1990) who distinguished between voiced and voiceless laughs. In their article Not All Laughs Are Alike, Bachorowski and Owren (2001) made a distinction between laughs that are song -like (voiced), snort -like (with nasal cavity turbulence) and grunt -like (with laryngeal and oral friction). Those classifications take phonetic dimensions of vocal production into account. Transcripts in normal writing usually do not contain non-verbal vocalizations such as laughter, noises of breathing or throat-clearing, which are hard to write down because of an unclear syllable structure and standardized orthography. Nevertheless, there are stylized forms of spelled laughter such as hehe. The idea behind haha, ahah and jaja is that a typical laugh is made up of 344

7 Laughter two syllables, with each syllable having a consonant and a vowel. The number of syllables in song-like laughs can of course change. What looks like vowel consonant alternations is a voiced voiceless pattern in an expiratory phase (episode). Inspiratory elements can intersperse the episodes (see, e.g., Trouvain, 2003; Chafe, 2007). In addition to freestanding laughter, we can observe that laughing can be combined with articulated speech. So-called speech-laughs (Nwokah et al., 1999) can sometimes count up to 50% of all laughs of a person. From a vocal production perspective, speech-laughs are characterized with a reinforced expiratory activity and a tremor-like voice quality while articulating (Trouvain, 2001). In scripted versions of interviews one can sometimes find the information when and at which time the interviewee has laughed by inserting meta-information by means of tags such as (laughs). The reader is referred to the laughable in the text but has no information about how this laugh was produced. Similarly, emoticons such as ;-) (= winking), : ) (= smiling), and LOL (= laughing out loud) can help to enrich and to clarify more informal texts in s, text messages, and chat contributions. The advantage of these emoticons is that the authors deliberately inserted these pieces of information in contrast to the transcriber of the interview who decides if this is additional information of someone else s spoken text production. Laughter in comics and cartoons can be recognized by the pictures, but additional iconic information can be delivered by changes in the typography of the transcribed laugh in the bubble. In discourse and conversation, analytical transcription systems such as the Gesprächsanalytische Transkriptionssystem 2 (GAT 2) (Selting et al., 2009, p. 367), it is recommended to transcribe laughter either as hahaha, hehe, hihi when it is a shorter or syllabic laugh corresponding to the number of pulses or laugh syllables, or as ((laughs)) between double rounded brackets as a description of laughter. Speech-laughs or laughed speech should be transcribed as <<laughing> word > whereas the convention for smiled speech is << :-) > word>. In their example transcript (1) taken from a real conversation, Selting et al. (2009, p. 394) depict laughter at various places. The transcriptions of various laughing events differ from event to event: In line 28 the laughs start with phhh, continued by hohoho; in lines 38 and 39 hm; in lines 40 and 41 it is lacht verstohlen, ca. 1.2 Sek lang; in lines 62, 63 and 65 it is hehehe or he he; and in line 66 he ho ho. (Interlinear translations of the text were inserted below the respective lines in the transcript.) (1) 27 S2: und sich mit den NACHbarn ange[legt,=ne, ] (1.2) and PRON with the neighbours argued PTCL and argued with the neighbours 28 S1: [phhh hohoho] S2: und ham wir immer gesagt HIT[ler;=ne,] and have we always said Hitler PTCL and we always said Hitler 37 S1: [HITler;] 38 S1: h[m,] 39 S2: [hm,] 40 S1: [((lacht verstohlen, ca. 1.2 Sek. lang))] [((laughs in a stealthy way, ca. 1.2 sec.))] 41 S2: [((lacht verstohlen, ca. 1.2 Sek. lang))]

8 Jürgen Trouvain and Khiet P. Truong 61 S2: jetzt wohnt er nämlich mit seiner SCHW:ESter now lives he actually with his sister zusa[mmen.] together now he actually lives together with his sister 62 S1: [hehehe] 63 S1: he [he] 64 S2: [und] die LAUfen: (.) RUM wie n URaltes and they run around like an old-age ehe[paar.]=ne, couple PTCL and they behave like like an old-age couple 65 S1: [he he] 66 S1: he ho [ho] 67 S2: [OUH] mann. (.) heh PTCL man Oh boy! The diversity of these transcriptions shows various interpretations of various laughter events, be it a laugh with a syllabic structure (hohoho), laughs with different vowel qualities (he vs. ho), laughing without any vowel articulation (hm), with a more literal description (verstohlen in German or stealthy in English) or with temporal information (ca. 1.2 Sek. lang). Especially with the interpretative comments, it is unclear whether a second transcriber would come up with the same or similar transcription of the laughing events. While it is not the aim of this chapter to criticize this kind of diversity in transcription, it shows two things: first, each transcription is always an interpretation. This holds in particular for non-words without standard spelling. And second, the diversity of the ways to transcribe laughter mirrors the diversity of laughter in general. The preceding transcript shows a further important feature of laughter in interaction: all laughs in this example transcript were overlapping, either with speech of the co-participant (e.g., lines 28 65) or with laughter of the co-participant (e.g., lines 38 41). It is not unusual that speakers overlap with their vocalizations. The conversational principle one speaker at a time (Sacks et al., 1974; Stivers, 2009) does not always hold (see, e.g., Liddicoat, 2007). This restriction is supported by corpora of conversational speech where a considerable amount of cross-talk was observed (e.g., Campbell, 2007; Heldner & Edlund, 2010). Laughter in particular has a tendency to overlap with laughter as could be shown by Laskowsi and Burger (2007), Truong and Trouvain (2012b), and also Smoski and Bachorowski (2003). Laughter seems to serve as the optimal opportunity for a joint vocalization. In lines 40 and 41 of the transcript we see synchronous laughter of both interlocutors, whereas the laughter in lines 28, 62, 63 and 65 overlaps with speech of the interlocutor. Interestingly, the overlaps in the latter cases are transcribed as aligned on a syllabic basis, e.g., he [he] in line 63 aligns with [und] in line 64. There is no standard annotation of non-verbal vocalizations in speech corpora. Usually transcriptions of laughter are made for the whole laughter bout on the word level. The following patterns can often be observed. A laugh is marked as a single element but with an extra mark such as angle brackets: <laugh>. A speech-laugh : that was <laugh: so funny>. Smiled speech: that was <smile: so funny>. Breathing noises are often <h>; however, it remains often unclear whether the breathing should be regarded as part of the laugh or not. 346

9 Laughter 2.3 Placement and Timing Details of Laughing It was stressed that laughter usually occurs in spoken interaction and that the social dimension is also reflected in common or shared laughter. Jefferson (1979, p. 93) describes the sharing of laughter with an invitation acceptance scheme : Laughter can be managed as a sequence in which speaker of an utterance invites recipient to laugh and recipient accepts that invitation. One technique for inviting laughter is the placement, by speaker, of a laugh just at completion of an utterance, and one technique for accepting that invitation is the placement, by recipient, of a laugh just after the onset of speaker s laughter. This normally leads to the overlap of laughs. The exact mechanisms of how interlocutors manage to produce an overlapping laugh in talk-in-interaction remain unclear. 2.4 Measuring Laughter It is possible to observe the vocal production of laughter of the phonetic levels of physiological generation, its acoustic characteristics, and its perception. In contrast to speech sounds, which are mainly the product of the control of articulatory gestures in the vocal tract, laughter mainly reflects modulations of phonatory and respiratory movements. This means that vowels that can be observed in laughter are not produced with a specific articulatory target (Bachorowski et al., 2001). Invasive physiological measurements of laughing with a focus on the respiratory apparatus or on glottal activities in the larynx (e.g., Luschei et al., 2006; Esling, 2007) have the advantage of observing laughter directly at the source of generation. The disadvantage is that the data is gained in settings that are far away from persons laughing in interaction and in an ideally unobserved way. Nevertheless, measurements under pure laboratory conditions can give important details of the mechanics of laughter, but should be backed with data recorded in more natural or naturalistic situations. Non-invasive kinematic studies of rib cage and abdominal movements in spontaneous dialogues show how tightly coupled the inhalation and exhalation activities of both conversational partners can be (McFarland, 2001). Though not every inhalation results in an audible noise, many song-like laughs have such an inbreath noise as an offset (Chafe, 2007). Sometimes they co-occur with linguistically motivated inhalation noises between major prosodic phrases, which usually reflects syntactical and discourse structure. Compared to speech, laughter can show dramatic differences with respect to prosodic dimensions of fundamental frequency (pitch) and intensity (loudness). It is the usual tendency for song-like laughs to be higher and louder than spoken language. There seems to be a difference between males and females: men produce much more unvoiced laughter than women (Bachorowski et al., 2001). Laughs are not alike this statement is valid for the different acoustic characteristics of laugh productions but also for the effect of laughter when perceiving it. Laughter is sometimes described as staccato-like, meaning that a laugh consists of a sequence of very similar laugh syllables. However, it was shown that a strict rhythmicity by repeating the same syllable has been perceived as less natural than introducing some variation in this rhythmic pattern (Kipper & Todt, 2001; 2003a; 2003b). With regard to the participation of the voice, it is interesting to see that unvoiced laughs are considered less positive than voiced forms (Cirillo & Todt, 2005). Variation in the prosodic parameter duration (length), fundamental frequency (pitch), and intensity 347

10 Jürgen Trouvain and Khiet P. Truong (loudness), as well as variation regarding voice quality (Esling, 2007), can have very different effects of which we have so far nearly no exact knowledge. Apart from quantitative and qualitative aspects, it seems to be the case that laughter as other affective triggers does influence perceptual dimensions such as the memory in the sense of laugh and you will be remembered (Armony et al., 2007). 3 Methodology and Sample Analysis In this section a selected sample of a joint laugh taken from a conversation is analyzed with a particular focus on the timing of the laughing interlocutors. First a transcript is provided, followed by an interpretation of the acoustic representation of the same sample. 3.1 Data The selected sample of an overlapping laugh is taken from a conversation in which interlocutors laugh together, because laughter seems to represent an optimal opportunity for joint vocalization, be it with speech or laughter of the interlocutor (Laskowski & Burger, 2007; Truong & Trouvain, 2012b; Smoski & Bachorowski, 2003). Exceptions of the principle one speaker a time are rather frequent in spontaneous speech for instance, as feedback expression or collaborative completions, which are continuations by the conversational partner with matching prosodic features (e.g., Lerner, 1991; Local, 2005). Other joint vocalizations that require acrossspeaker coordination are, for instance, synchronous reading aloud (Cummins, 2007), singing in a choir, and a common prayer in church. The examples shown in this section stem from a dialogue from the Lindenstraße Corpus (IPDS, 2006), where both conversational partners had to talk about a German TV soap (before the recording they had to order non-identical video clips of a given episode). Both interlocutors (same sex, friends) wore head-mounted microphones and earphones and were seated in different sound-treated rooms. The room separation has the advantage to suppress the visual information exchange and to provide acoustic recordings with separated acoustic channels (no acoustic mixture). More details of the analysis of these examples can be found in Trouvain and Truong (2014). 3.2 Transcript The transcript of the example in (2) shows the spoken words and non-verbal vocalizations for a speaker s turn. Each line has been translated into English in an interlinear way; a summary of each turn is printed in a smaller font at the bottom. Speaker L produces a speech-laugh during the articulation of machen die leut da. She pauses (indicated by dashes) before she audibly inhales (indicated by h). The square brackets for both speakers mark the overlapping part. Speaker S joins in the speech-laugh of speaker L. Her laughing (consisting of seven syllables plus inhalation) co-occurs with the rest of L s laughter before S is taking over the turn. This pattern corresponds to the invitation acceptance scheme described earlier. The laughable part seems to refer to the reported speech mai, was pfn machen die leut da. Laughable elements could be mocking the character they are talking about, which may include the dialect imitated here. Also, the speech error marked by the interjection could be interpreted as laughable as well as the content of what was quoted. There are several possibilities of interpretations. Speaker S changes the communicative role with her overlapping laughter, before the laugh listener, after the laugh speaker. 348

11 Laughter (2) L: is halt son skelett im flur und sie meint dann so is just such a skeleton in corridor and she says then so mai, was <<interjection> pfn> well what INTERJ <<laughing>machen die leut [da> (---) h] make the people there There is a skeleton in the corridor and she goes like "well, what are the people doing there?" S: [hehehehehehehe h] echt nee hier wars so, dass wir ne kurze szene wo sie really no here was so that we a short scene where she in der straße entlanggehen und dann h in the street go along and then Really? No, here we had a short scene where she walked along the street and then Acoustic Representation Figure 24.1 illustrates the example section from transcript (2). Recordings with separate channels allow us to determine the exact timing of the laughter in the dialogue flow and the exact phonetic-prosodic realizations of the laughter by each speaker (top and middle graph). In recordings without channel separation (bottom part of Figure 24.1) these details of prosody are masked. Figure 24.1 Example of laughter at turn-taking, from Lindenstraße Corpus (IPDS, 2006), (Dialogue 4, from 164 to 173 sec) represented as wave form and spectrogram (0 8 khz). Top: left channel speaker. Mid: right channel speaker. Bottom: mixed speech signal. Marked with shading: the laugh production 349

12 Jürgen Trouvain and Khiet P. Truong Figure 24.1 (Continued) Mixed signals seem to be the case in most dialogue corpora investigated in interactional linguistics. In addition, some less intense vocalizations may not even be noticeable when masked in a recording without channel separation. The exact timing for both speakers in Figure 24.1 shows that speaker S (middle) stops laughing and starts speaking before speaker L (top) terminates her laugh. In contrast to the transcript 350

13 Laughter where both speakers ended their laughs synchronically, the analysis of the separated recordings makes visible and audible the more fine-grained mechanisms of laughing and laughing together. It is not a coincidence that the turn-giving speaker L starts a bit later with her clearly audible inhalation than the turn-taking speaker S. 4 New Debates There are plenty of unsolved issues regarding laughter research and some of them seem to be untouched. Laughter is a multidisciplinary research object. As a consequence, different communities such as psychologists, evolutionary biologists, speech technologists, phoneticians, and interactional linguists discuss relevant topics often independently from each other at conferences and journals. Natural hot topics include the contribution of laughter to affectivity and emotions, and the still unsolved puzzle of how laughing and smiling are linked in everyday communication. Interesting fields involve how deaf people use laughter and also the topic of laughter-like vocalizations produced by non-human species. Limitations in space reduce this section to the presentation of a few further topics for future research. 4.1 Production and Perception of Laughter Vocalizations From a phonetic point of view it is still underexplored how the respiratory, the glottal and the vocal tract activities interact for different types of laughter. The control of the vocal apparatus while laughing is not only a mechanical manoeuvre at different places but also requires neurological routines that allow quick reactions in interactional laughing. Individual differences in laughter production are often mentioned but infrequently investigated. Largely unexplored is how we perceive laughter. This topic also includes the visual aspects of laughing. Often we use contextual information to interpret the laughing signal. Thus, from an interactional linguistic point of view, it is interesting to learn more about the pragmatic functions laughter can have in spontaneous discourse. 4.2 Interpretation of Laughter and Its Cultural Display Rules Probably the most important question each perceived laugh evokes is is the other laughing with me or at me? Laughter can be used for social inclusion as well as social exclusion. Particularly, the study of laughter that leads to exclusion, either being intended by the speakers or being interpreted by the perceivers, is a fruitful field for further research for instance, in gelotophobia (the fear of being laughed at), which is a special type of social phobia (see, e.g., Ruch & Proyer, 2008). Another important group with social challenges with respect to laughter is people with an autistic spectrum disorder (e.g., Hudenko et al., 2009). Similar to other non-verbal vocalizations like sneezing, yawning, and coughing, laughter is under cultural display rules. Laughter as all kind of behavior always happens in a social framework in which people consider some degree of laughter as inacceptable, inappropriate, or otherwise outside an assumed norm. Our knowledge of what is usually considered laughable and what kind of laughter and the timing of laughter are considered appropriate is rather limited. Nevertheless, we are aware of groups of people violating the expected ranges, often combined with affective-social problems or neurological disorders such as aphasia (Rohrer et al., 2009). Cultural settings often provide regulations whether and how laughter is displayed in given situations. In Christian traditions smiling and laughter are usually not displayed (e.g., in displays of Jesus), in contrast to, for instance, Buddhist traditions. Sometimes, smiling and laughter are used 351

14 Jürgen Trouvain and Khiet P. Truong to signal two different categories of behavior. Artwork of the Last Judgement shows people with a smile (like angels) going to heaven whereas their peers with a laugh (like madness) going to hell. In visualizations of laughter in historic illustrations, Hoffmann et al. (2012) analyzed the facial features of schadenfreude laughter by applying the Facial Action Coding System. Results show that while the encoding of schadenfreude laughter is heterogeneous, schadenfreude is decoded when the facial expression unites markers of joy (orbicularis oculi, or pars orbitalis, muscle and the zygomatic major muscle), as well as markers of negative emotions (e.g., brow lowering), or in one case, where the initially categorized schadenfreude illustration contained markers distorting the expression of joy (e.g., frowning and the lowering of lip corners). These findings support the hypothesis that schadenfreude may be expressed by a morphologically distinct blend of a positive and a negative emotion, or is expressed by joyful laughter (with the expression being modulated due to social desirability). 4.3 Technological Applications With Laughter Vocalizations Laughter in dialogues, particularly those with more than two interlocutors, can represent important aspects in automatic speech recognition. First, laughter as a frequently occurring non-word vocalization can complicate the recognition result an issue that should be avoided by a better recognition of laughing events in speech. Second, automatic laughter detection in dialogues can help to reconstruct discourse acts like topic change. A different technological application of laughter is to enrich the expressivity of synthetic speech (Campbell, 2006), especially the expressivity of personalized synthetic voices. Speech synthesis is a great communicative help for people who are unable to articulate speech. However, the expressiveness of the person who is reliant on a speech synthesizer is clearly reduced when there is no possibility to show laughter, or a poor realization of appropriate laughs. From a perceptual perspective, another field of technology could be helpful for example, the development of computer-assisted training material for autistic persons and other people who have problems recognizing and/or interpreting certain types of laughter (e.g., Tanaka & Campbell, 2014). Taken together, there are many technological applications conceivable where laughter can play an important role in improving the communication of machines with human users. However, in order to develop appropriate and suitable laughter interactive technology, we must first gain a good understanding of what laughter is. Note 1 Thanks to John Laver for this list. References Adelsward, V. (1989). Laughter and dialogue: The social significance of laughter in institutional discourse. Nordic Journal of Linguistics, 12, Armony, J. L., Chochol, C., Fecteau, S., & Belin, P. (2007). Laugh (or cry) and you will be remembered: Influence of emotional expression on memory for vocalizations. Psychological Science, 18(12), Bachorowski, J. A., & Owren, M. J. (2001). Not all laughs are alike: Voiced but not unvoiced laughter elicits positive affect in listeners. Psychological Science, 12, Bachorowski, J. A., Smoski, M. J., & Owren, M. J. (2001). The acoustic features of human laughter. Journal of the Acoustical Society of America, 111(3), Belin, P., Fillion-Bilodeau, S., & Gosselin, F. (2008). The Montreal affective voices: A validated set of nonverbal affect bursts for research on auditory affective processing. Behavior Research Methods, 40(2),

15 Laughter Campbell, N. (2006). Conversational speech synthesis and the need for some laughter. IEEE Transactions on Audio, Speech, and Language Processing, 14(4), Campbell, N. (2007). Approaches to conversational speech rhythm: Speech activity in two-person telephone dialogues. In Proceedings from the 16th International Congress of Phonetic Sciences (pp ). Saarbrücken, Germany. Chafe, W. (2007). The importance of not being earnest. Amsterdam, Netherlands: Benjamins. Cirillo, J., & Todt, D. (2005). Perception and judgement of whispered vocalisations. Behaviour, 142, Cummins, F. (2007). Speech synchronization: Investigating the links between perception and action in speech production. In Proceedings from the 16th International Congress of Phonetic Sciences (pp ). Saarbrücken, Germany. Darwin, C. (1872/1998). The expression of the emotions in man and animals (3rd ed., P. Ekman, Ed.). London: Oxford University Press. Davila-Ross, M., Hutchinson, J., Russell, J. L., Schaeffer, J., Billard, A., Hopkins, W. D., & Bard, K. A. (2014). Triggering social interactions: Chimpanzees respond to imitation by a humanoid robot and request responses from it. Animal Cognition, 17, Davila-Ross, M., Owren, M. J., & Zimmermann, E. (2009). Reconstructing the evolution of laughter in great apes and humans. Current Biology, 19, Ekman, P., & Friesen, W. V. (1982). Felt, false, and miserable smiles. Journal of Nonverbal Behavior, 6, Esling, J. (2007). States of the larynx in laughter. In Proceedings from The Interdisciplinary Workshop on The Phonetics of Laughter (pp ). Saarbrücken, Germany. Glenn, P. (1989). Initiating shared laughter in multi-party conversations. Western Journal of Speech Communication, 53, Glenn, P. (2003). Laughter in interaction. Cambridge: Cambridge University Press. Glenn, P., & Holt, E. (Eds.). (2013). Studies of laughter in interaction. London: Bloomsbury. Grammer, K., & Eibl-Eibesfeldt, I. (1990). The ritualisation of laughter. In W. A. Koch (Ed.), Natürlichkeit der Sprache und der Kultur: acta colloquii (pp ). Bochum, Germany: Brockmeyer. Habermann, G. (1955). Physiologie und phonetik des lauthaften lachens. Leipzig, Germany: J. A. Barth. Heldner, M., & Edlund, J. (2010). Pauses, gaps and overlaps in conversations. Journal of Phonetics, 38, Hoffmann, J., Ruch, W., & Platt, T. (2012). The en- and decoding of schadenfreude laughter sheer joy expressed by a Duchenne laugh or emotional blend with distinct morphological expression? Proceeding from 3rd Workshop on Laughter and Other Non-Verbal Vocalisations, Dublin, Ireland. Holt, E. (2010). The last laugh: Shared laughter and topic termination. Journal of Pragmatics, 42, Hudenko, W. J., Stone, W., & Bachorowski, J. A. (2009). Laughter differs in children with autism: An acoustic analysis of laughs produced by children with and without the disorder. Journal of Autism and Developmental Disorders, 39(10), IPDS. (2006). The Kiel Corpus of Spontaneous Speech, Volume 4 [DVD]. Kiel: IPDS. Jefferson, G. (1979). A technique for inviting laughter and its subsequent acceptance/declination. In G. Psathas (Ed.), Everyday language: Studies in ethnomethodology (pp ). New York, NY: Irvington. Jefferson, G., Sacks, H., & Schegloff, E. (1987). Notes on laughter in the pursuit of intimacy. In G. Button & J. Lee (Eds.), Talk and social organisation (pp ). Clevedon: Multilingual Matters. Kipper, S., & Todt, D. (2001). Variation of sound parameters affects the evaluation of human laughter. Behaviour, 138, Kipper, S., & Todt, D. (2003a). Dynamic-acoustic variation causes differences in evaluations of laughter. Perceptual and Motor Skills, 96, Kipper, S., & Todt, D. (2003b). The role of rhythm and pitch in the evaluation of human laughter. Journal of Nonverbal Behaviour, 27, Krys, K., et al. (2016). Be careful where you smile: Culture shapes judgments of intelligence and honesty of smiling individuals. Journal of Nonverbal Behaviour, 40, Labov, W. (1972). Sociolinguistic patterns. Philadelphia: University of Pennsylvania. Laskowski, K., & Burger, S. (2007). Analysis of the occurrence of laughter in meetings. In Proceedings from the Interspeech Conference (pp ). Antwerp, Belgium. Lerner, G. H. (1991). On the syntax of sentences in progress. Language in society, 20, Liddicoat, A. J. (2007). An introduction to conversation analysis. New York, NY: Continuum. Local, J. (2005). On the interactional and phonetic design of collaborative completions. In W. Hardcastle & J. M. Beck (Eds.). A figure of speech: A festschrift for John Laver (pp ). Mahwah, NJ: Lawrence Erlbaum. 353

16 Jürgen Trouvain and Khiet P. Truong Luschei, E. S., Ramig, L. O., Finnegan, E. M., Baker, K. M., & Smith, M. E. (2006). Patterns of laryngeal electromyography and the activity of the respiratory system during spontaneous laughter. Journal of Neurophysiology, 96(1), Makagon, M. M., Funayama, E. S., & Owren, M. J. (2008). An acoustic analysis of laughter produced by congenitally deaf and normally hearing college students. Journal of the Acoustical Society of America, 124, McFarland, D. H. (2001). Respiratory markers of conversational interaction. Journal of Speech, Language and Hearing Research, 44(1), Mehu, M. (2011). Smiling and laughter in naturally occurring dyadic interactions: Relationship to conversation, body contacts, and displacement activities. Human Ethology Bulletin, 26, Mehu, M., & Dunbar, R.I.M. (2008). Naturalistic observations of smiling and laughter in human group interactions. Behaviour, 145, Nwokah, E. E., Hsu, H. C., Davies, P., & Fogel, A. (1999). The integration of laughter and speech in vocal communication: A dynamic systems perspective. Journal of Speech, Language and Hearing Research, 42, O Donnell-Trujillo, N., & Adams, K. (1983). Heheh in conversation: Some coordinating accomplishments of laughter. Western Journal of Speech Communication, 47, Provine, R. R. (2000). Laughter: A scientific investigation. London: Faber & Faber. Rohrer, J. D., Warren, J. D., & Rossor, M. N. (2009). Abnormal laughter-like vocalisations replacing speech in primary progressive aphasia. Journal of the Neurological Sciences, 284, Ruch, W., & Proyer, R. T. (2008). The fear of being laughed at: Individual and group differences in gelotophobia. Humor: International Journal of Humor Research, 21, Sacks, H. (1992). Lectures on conversation (Vols. 1 2, G. Jefferson, Ed.). Oxford: Blackwell. Sacks, H., Schegloff, E. A., & Jefferson, G. (1974). A simplest systematics for the organisation of turn-taking for conversation. Language, 50, Sauter, D. A. (2010). More than happy: The need for disentangling positive emotions. Current Directions in Psychological Science, 19, Sauter, D. A., Eisner, F., Ekman, P., & Scott, S. K. (2010). Cross-cultural recognition of basic emotions through nonverbal emotional vocalizations. PNAS Proceedings of the National Academy of Sciences, 107(6), Schenkein, J. N. (1972). Towards the analysis of natural conversation and the sense of heheh. Semiotica, 6, Selting, M., et al. (2009). Gesprächsanalytisches Transkriptionssystem 2 (GAT 2). Gesprächsforschung Online-Zeitschrift zur verbalen Interaktion, 10, Smoski, M. J., & Bachorowski, J. A. (2003). Antiphonal laughter between friends and strangers. Cognition and Emotion, 17(2), Stivers, T. (2009). Universals and cultural variation in turn-taking in conversation. Proceedings of the National Academy of Sciences of the United States of America (PNAS), 106(26), Szameitat, D. P., Alter. K., Szameitat, A. J., Dietrich, S., Wildgruber, D., Sterr, A., & Darwin, C. J. (2009). Acoustic profiles of distinct emotional expression in laughter. Journal of the Acoustical Society of America, 126, Tanaka, H., & Campbell, N. (2014). Classification of social laughter in natural conversational speech. Computer Speech and Language, 28(1), Thibault, P., Levesque, M., Gosselin, P., & Hess, U. (2012). The Duchenne marker is not a universal signal of smile authenticity but it can be learned! Social Psychology, 43, Trouvain, J. (2001). Phonetic aspects of speech-laughs. In Proceedings from the 2nd Conference on Orality & Gestuality (ORAGE) (pp ) Aix-en-Provence, France. Trouvain, J. (2003). Segmenting phonetic units in laughter. In Proceedings from 15th International Conference of the Phonetic Sciences (ICPhS) (pp ) Barcelona, Spain. Trouvain, J. (2014). Laughing, breathing clicking The prosody of nonverbal vocalisations. In Proceedings from Speech Prosody (SP7) (pp ). Dublin, Ireland. Trouvain, J., & Campbell, N. (Eds.). (2007). Proceedings from the International Workshop on The Phonetics of Laughing. Saarbrücken, Germany. Trouvain, J., & Truong, K. P. (2014). Towards unravelling prosodic characteristics of speaker-overlapping laughing in conversational speech corpora. In D. Barth-Weingarten & B. S. Reed (Eds.), Prosodie und Phonetik in der Interaktion Prosody and phonetics in interaction (pp ). Mannheim, Germany: Verlag für Gesprächsforschung. 354