Social Items



Anyone interested in the study of humor is faced with a terminological problem, which may appear trivial at first, but that eventually becomes serious enough to require clarification. Humor research has standardized on the use of the umbrella term ‘humor’ to indicate any form of communicative behavior intended or interpreted as having the intention to elicit amusement, mirth, laughter, or associated feelings of exhilaration, the perception of the comical and similar states of mind. By definition,‘humor’ is meant to encompass any form of such behavior, without any attempt at further differentiation. Under this term, humor encompasses most uses of irony. This is not to say that one cannot establish internal operational subdivisions, and study wit in 18th century England, for example. The same goes for irony and sarcasm: there are those who claim that the two are distinct phenomena, but the general consensus is that sarcasm is an aggressive form of irony.1 We generally do not differentiate between the two, as it is usually impossible to do so reliably in the intermediate cases. However, the difference may be reintroduced in a methodologically controlled way: for example, Cheang and Pell (this book) and Caucci and Kreuz (2012) examine specifically critical irony (sarcasm) as distinguished from “positive/humorous” (i.e., non critical) irony.


2. The state of the art
We can start with the observation that very little research has concerned itself in general with the prosody of humor. The available research is summarized and reviewed in Pickering et al. (2009). Conversely, there is a significant amount of literature on the prosody of irony. The central thrust of this research can be summed up as attempts to describe the “ironical tone of voice”. More broadly, researchers try to describe how irony is “marked” including non-prosodic and even non-linguistic markers. We will briefly review the state-of-the-art of the study of prosody and humor, laughter, and then turn to irony.


2.1 Humor
Pickering et al. (2009) examined jokes (short humorous narratives ending in a punch line) and found that the punch lines were not marked prosodically — neither by changes in pitch, volume, or speech rate, nor by significant pauses.
This negative finding is broadened to conversational (non-narrative) humor in Bertrand and Priego-Valverde, in Attardo et al., in Flamson et al., and to professional actors (Urios-Aparisi and Wagner). Other forms of marking (smiling, laughter) were found, but they are inconsistent.

2.2 Laughter
Laughter has attracted a significant body of work since the early work of Mowrer et al. (1987). In particular the work of Provine (1991, 1993, 1996a, b, 2000, 2004a, b; on Provine’s work, see Ruch 2002), of Bachorowski and her associates (e,g., Bachorowski and Owren 2001; Bachorowski et al. 2001; Owren and Bachorowski 2003; Smoski and Bachorowski 2003), Trouvain (2001, 2003; Trouvain and Schroder 2004), and Vettin and Todt (2004, 2005), Kipper and Todt (2005), Gervais and Wilson (2005), O’Connell and Kowal (2005, 2006) stand out, but Chafe (2007) is the most comprehensive work on the subject. This is not the place to review this literature, but let us note an aspect relevant to the articles in this issue: Provine’s claim (2000) that laughter punctuates speech, i.e., does not occur within words or phrases, has been refuted (Nwokah et al. 1999; Chafe 2007).


2.3 Irony
We will primarily concern ourselves with phonological (and specifically pitch) and facial markers. There exist morphological, syntactic, lexical (e,g., Caucci and Kreuz 2012), and typographical markers as well (see Haiman 1998: 28–60; Attardo 2000), but we will not address those in this context. It should also be noted that the existence of an unmarked (deadpan) delivery is commonly assumed. Hancock (2004) reports that 76% of ironical utterances are marked, i.e., about 24% are produced with deadpan delivery.


2.4 Phonological markers
The literature on the markers of irony and sarcasm includes several studies on phonological markers of sarcasm.
The most frequently mentioned are discussed below. The most commonly noted index of ironical intent is intonation. The ironical intonation has been described as a flat (neither rising nor falling) contour (Milosky and Wrobleski 1994; Shapely 1987; Fonagy 1975; Myers Roy 1978: 58, quoted in Barbe 1995; Haiman 1998: 35–36). Schaffer (1982: 45) reports question intonation (i.e., rising) as a marker of irony.
Anolli et al. (2000) found that lower pitch indicated irony. Rockwell (2000) also found that a lower pitch was a marker of irony, among trained performers. Similarly, Haiman discusses “inverse pitch obtrusion” (i.e., the utterance of the stressed syllable “at a lower pitch than the surrounding material” 1998: 31) in English and German.
Muecke (1978: 370–371), Schaffer (1981), Adachi (1996), and Haiman (1998) report that an exaggerated pitch or extremes of pitch mark irony. Similarly, Attardo et al. (2003) claim that “exaggerated pitch marks sarcasm”.
Haiman (1998: 30–41) discusses several other intonational patterns that can be used to indicate sarcasm: such as singsong melody, falsetto, “heavy exaggerated stress and relatively monotonous intonation” (Haiman 1998: 39) and separation by “heavy” (i.e., long) pauses of the words (Schaffer 1982: 45; Haiman 1998: 39, for Japanese and German). Muecke (1978: 370) reports the use of “softened voice’’.
Bolinger (1985: 127, 1989: 75–76) points to the use of rise-fall contours with ironical statements such as “is that so” or “you don’t say”, and low tones with statements such as “a likely story” or “I’ll bet”. Caucci and Kreuz (2012) also note that some lexical expressions seem to connote “sarcasm”.
The use of a marked succession of prominent syllables is analyzed as “beat clash” by Uhmann (1996), and is argued to provide a cue to irony. Stress patterns broader than usual are also reported by several authors: Cutler (1974: 117), Myers Roy (1977: 58, quoted in Barbe 1995), Schaffer (1982: 45), and Barbe (1995: 76).
Several authors report that nasalization is a marker of ironical intent, e.g., Cutler (1974: 117), Muecke (1978: 370, who employs the term ‘mycterism’), Myers Roy (1977: 58, quoted in Barbe 1995), Schaffer (1982: 45), Chen (1990: 28), and Haiman (1998: 30–31).
Speech rate may also be a factor, with Cutler (1974: 117) and Fonagy (1971: 42) suggesting that a slowed speech rate may be indicative of irony and several authors pointing to syllable lengthening as a possible cue (Myers Roy 1977: 58, quoted in Barbe 1995; Schaffer 1982: 45; Haiman 1998: 34 in Chinese and several other languages; Adachi 1996: 8, for Japanese). Laughter syllables scattered in the utterance or preceding or following it have also been reported as markers of irony (Schaffer 1982: 45; Haiman 1998: 31). The literature on the use of laughter to mark humorous (in general) intention on the speaker’s part is ample (see e.g., Jefferson 1984, 1985).


2.5 Facial markers
Among the facial signals of ironical intent the following have been quoted in the literature:
-          Eyebrows: raised, lowered
-          Eyes: wide open, squinting, rolling
-          Winking (Muecke 1978: 368–369)
-          Nodding (Caucci and Kreuz 2012)
-          Lip tightening (Ibid.)
-          Smiling
-          Blank face (Attardo et al. 2003)
-          Prosody-face incongruity (Pell 2005: 211)
-          Gaze aversion: Williams et al. (2009) show that gaze aversion signals irony.
This is not an upward “rolling of the eyes” (Attardo et al. 2003), but a sideways (horizontal) movement. In some case it was downward. Interestingly, Williams et al. find that gaze aversion was a decrease on the speaker’s baseline, not an absolute factor. It is unclear whether speakers are intentionally signaling their ironical intention or whether this is a case of what Ekman and Friesen (1969) would call a “leak” (i.e., an involuntary revealing of one’s mental state).

2.6 Conclusions of the above discussions
One definite conclusion that can be drawn from these studies is that the proper mode of investigation for humor, irony, sarcasm, and related subjects is multimodality. Both Caucci and Kreuz (2012) and Attardo et al. (this book) stress the importance of multimodal analyses and in particular the significance of smiling as a marker of humor/irony. The interplay of visual/facial factors, intonational factors, and semantic and contextual factors is a wide open field that is obviously in need of much further research, as Attardo et al. (2003) already noted. The multimodality of humor can be connected to underlying cognitive processes. As Wennerstrom shows, a cognitive process such as “blending” underlies the prosody of “intonation jokes” and adds a humorous component to the lexico-grammatical structure of the jokes. Such cognitive processes are involved in other meaning-making activities such as metaphor that also use multimodal resources (see Forceville and Urios- Aparisi 2009).
We can state with a certain degree of confidence that the picture of whether and how irony is marked prosodically and behaviorally is complex — there is no easy one-to-one mapping, no univocal “ironical tone of voice”. There may be only differential markings as Bryant and Fox-Tree (2005) and Attardo et al. (2003) have argued, or substantive markers may be associated with the aggressive aspect of sarcasm, as Bryant (this book) concludes. However, recent research using ERP (Event Related Potentials), such as electroencephalography, has shown an interesting lack of correlation between ironic prosody and its perception (Regel et al. 2006; Amenta and Balconi 2008).
There seem to be definite differences across languages: see Adachi (1996) for Japanese, Anolli et al. (2002) for Italian, Laval and Bert-Erboul (2005), Bertrand and Priego-Valverde (this book) for French, and Cheang and Pell (this book), for Cantonese Chinese. There are also differences across age groups: Capelli et al. (1990), Milosky and Ford (1997), and Laval and Bert-Eboul (2005) note that younger children rely more on prosodic cues. Whalen and Pexman (2010: 379) report that older children mode adopt (i.e., respond with irony to irony) more frequently, using prosodic and behavior markers. See also Meng (2006) for Chinese children.

By:

Salvatore Attardo
Texas A&M University-Commerce

Manuela Maria Wagner
University of Connecticut

Eduardo Urios-Aparisi
University of Connecticut

PROSODY AND HUMOR



Anyone interested in the study of humor is faced with a terminological problem, which may appear trivial at first, but that eventually becomes serious enough to require clarification. Humor research has standardized on the use of the umbrella term ‘humor’ to indicate any form of communicative behavior intended or interpreted as having the intention to elicit amusement, mirth, laughter, or associated feelings of exhilaration, the perception of the comical and similar states of mind. By definition,‘humor’ is meant to encompass any form of such behavior, without any attempt at further differentiation. Under this term, humor encompasses most uses of irony. This is not to say that one cannot establish internal operational subdivisions, and study wit in 18th century England, for example. The same goes for irony and sarcasm: there are those who claim that the two are distinct phenomena, but the general consensus is that sarcasm is an aggressive form of irony.1 We generally do not differentiate between the two, as it is usually impossible to do so reliably in the intermediate cases. However, the difference may be reintroduced in a methodologically controlled way: for example, Cheang and Pell (this book) and Caucci and Kreuz (2012) examine specifically critical irony (sarcasm) as distinguished from “positive/humorous” (i.e., non critical) irony.


2. The state of the art
We can start with the observation that very little research has concerned itself in general with the prosody of humor. The available research is summarized and reviewed in Pickering et al. (2009). Conversely, there is a significant amount of literature on the prosody of irony. The central thrust of this research can be summed up as attempts to describe the “ironical tone of voice”. More broadly, researchers try to describe how irony is “marked” including non-prosodic and even non-linguistic markers. We will briefly review the state-of-the-art of the study of prosody and humor, laughter, and then turn to irony.


2.1 Humor
Pickering et al. (2009) examined jokes (short humorous narratives ending in a punch line) and found that the punch lines were not marked prosodically — neither by changes in pitch, volume, or speech rate, nor by significant pauses.
This negative finding is broadened to conversational (non-narrative) humor in Bertrand and Priego-Valverde, in Attardo et al., in Flamson et al., and to professional actors (Urios-Aparisi and Wagner). Other forms of marking (smiling, laughter) were found, but they are inconsistent.

2.2 Laughter
Laughter has attracted a significant body of work since the early work of Mowrer et al. (1987). In particular the work of Provine (1991, 1993, 1996a, b, 2000, 2004a, b; on Provine’s work, see Ruch 2002), of Bachorowski and her associates (e,g., Bachorowski and Owren 2001; Bachorowski et al. 2001; Owren and Bachorowski 2003; Smoski and Bachorowski 2003), Trouvain (2001, 2003; Trouvain and Schroder 2004), and Vettin and Todt (2004, 2005), Kipper and Todt (2005), Gervais and Wilson (2005), O’Connell and Kowal (2005, 2006) stand out, but Chafe (2007) is the most comprehensive work on the subject. This is not the place to review this literature, but let us note an aspect relevant to the articles in this issue: Provine’s claim (2000) that laughter punctuates speech, i.e., does not occur within words or phrases, has been refuted (Nwokah et al. 1999; Chafe 2007).


2.3 Irony
We will primarily concern ourselves with phonological (and specifically pitch) and facial markers. There exist morphological, syntactic, lexical (e,g., Caucci and Kreuz 2012), and typographical markers as well (see Haiman 1998: 28–60; Attardo 2000), but we will not address those in this context. It should also be noted that the existence of an unmarked (deadpan) delivery is commonly assumed. Hancock (2004) reports that 76% of ironical utterances are marked, i.e., about 24% are produced with deadpan delivery.


2.4 Phonological markers
The literature on the markers of irony and sarcasm includes several studies on phonological markers of sarcasm.
The most frequently mentioned are discussed below. The most commonly noted index of ironical intent is intonation. The ironical intonation has been described as a flat (neither rising nor falling) contour (Milosky and Wrobleski 1994; Shapely 1987; Fonagy 1975; Myers Roy 1978: 58, quoted in Barbe 1995; Haiman 1998: 35–36). Schaffer (1982: 45) reports question intonation (i.e., rising) as a marker of irony.
Anolli et al. (2000) found that lower pitch indicated irony. Rockwell (2000) also found that a lower pitch was a marker of irony, among trained performers. Similarly, Haiman discusses “inverse pitch obtrusion” (i.e., the utterance of the stressed syllable “at a lower pitch than the surrounding material” 1998: 31) in English and German.
Muecke (1978: 370–371), Schaffer (1981), Adachi (1996), and Haiman (1998) report that an exaggerated pitch or extremes of pitch mark irony. Similarly, Attardo et al. (2003) claim that “exaggerated pitch marks sarcasm”.
Haiman (1998: 30–41) discusses several other intonational patterns that can be used to indicate sarcasm: such as singsong melody, falsetto, “heavy exaggerated stress and relatively monotonous intonation” (Haiman 1998: 39) and separation by “heavy” (i.e., long) pauses of the words (Schaffer 1982: 45; Haiman 1998: 39, for Japanese and German). Muecke (1978: 370) reports the use of “softened voice’’.
Bolinger (1985: 127, 1989: 75–76) points to the use of rise-fall contours with ironical statements such as “is that so” or “you don’t say”, and low tones with statements such as “a likely story” or “I’ll bet”. Caucci and Kreuz (2012) also note that some lexical expressions seem to connote “sarcasm”.
The use of a marked succession of prominent syllables is analyzed as “beat clash” by Uhmann (1996), and is argued to provide a cue to irony. Stress patterns broader than usual are also reported by several authors: Cutler (1974: 117), Myers Roy (1977: 58, quoted in Barbe 1995), Schaffer (1982: 45), and Barbe (1995: 76).
Several authors report that nasalization is a marker of ironical intent, e.g., Cutler (1974: 117), Muecke (1978: 370, who employs the term ‘mycterism’), Myers Roy (1977: 58, quoted in Barbe 1995), Schaffer (1982: 45), Chen (1990: 28), and Haiman (1998: 30–31).
Speech rate may also be a factor, with Cutler (1974: 117) and Fonagy (1971: 42) suggesting that a slowed speech rate may be indicative of irony and several authors pointing to syllable lengthening as a possible cue (Myers Roy 1977: 58, quoted in Barbe 1995; Schaffer 1982: 45; Haiman 1998: 34 in Chinese and several other languages; Adachi 1996: 8, for Japanese). Laughter syllables scattered in the utterance or preceding or following it have also been reported as markers of irony (Schaffer 1982: 45; Haiman 1998: 31). The literature on the use of laughter to mark humorous (in general) intention on the speaker’s part is ample (see e.g., Jefferson 1984, 1985).


2.5 Facial markers
Among the facial signals of ironical intent the following have been quoted in the literature:
-          Eyebrows: raised, lowered
-          Eyes: wide open, squinting, rolling
-          Winking (Muecke 1978: 368–369)
-          Nodding (Caucci and Kreuz 2012)
-          Lip tightening (Ibid.)
-          Smiling
-          Blank face (Attardo et al. 2003)
-          Prosody-face incongruity (Pell 2005: 211)
-          Gaze aversion: Williams et al. (2009) show that gaze aversion signals irony.
This is not an upward “rolling of the eyes” (Attardo et al. 2003), but a sideways (horizontal) movement. In some case it was downward. Interestingly, Williams et al. find that gaze aversion was a decrease on the speaker’s baseline, not an absolute factor. It is unclear whether speakers are intentionally signaling their ironical intention or whether this is a case of what Ekman and Friesen (1969) would call a “leak” (i.e., an involuntary revealing of one’s mental state).

2.6 Conclusions of the above discussions
One definite conclusion that can be drawn from these studies is that the proper mode of investigation for humor, irony, sarcasm, and related subjects is multimodality. Both Caucci and Kreuz (2012) and Attardo et al. (this book) stress the importance of multimodal analyses and in particular the significance of smiling as a marker of humor/irony. The interplay of visual/facial factors, intonational factors, and semantic and contextual factors is a wide open field that is obviously in need of much further research, as Attardo et al. (2003) already noted. The multimodality of humor can be connected to underlying cognitive processes. As Wennerstrom shows, a cognitive process such as “blending” underlies the prosody of “intonation jokes” and adds a humorous component to the lexico-grammatical structure of the jokes. Such cognitive processes are involved in other meaning-making activities such as metaphor that also use multimodal resources (see Forceville and Urios- Aparisi 2009).
We can state with a certain degree of confidence that the picture of whether and how irony is marked prosodically and behaviorally is complex — there is no easy one-to-one mapping, no univocal “ironical tone of voice”. There may be only differential markings as Bryant and Fox-Tree (2005) and Attardo et al. (2003) have argued, or substantive markers may be associated with the aggressive aspect of sarcasm, as Bryant (this book) concludes. However, recent research using ERP (Event Related Potentials), such as electroencephalography, has shown an interesting lack of correlation between ironic prosody and its perception (Regel et al. 2006; Amenta and Balconi 2008).
There seem to be definite differences across languages: see Adachi (1996) for Japanese, Anolli et al. (2002) for Italian, Laval and Bert-Erboul (2005), Bertrand and Priego-Valverde (this book) for French, and Cheang and Pell (this book), for Cantonese Chinese. There are also differences across age groups: Capelli et al. (1990), Milosky and Ford (1997), and Laval and Bert-Eboul (2005) note that younger children rely more on prosodic cues. Whalen and Pexman (2010: 379) report that older children mode adopt (i.e., respond with irony to irony) more frequently, using prosodic and behavior markers. See also Meng (2006) for Chinese children.

By:

Salvatore Attardo
Texas A&M University-Commerce

Manuela Maria Wagner
University of Connecticut

Eduardo Urios-Aparisi
University of Connecticut
Load Comments

Subscribe Our Newsletter

close