top of page

Humor Failure: A Review of Reasons and Explanations 

by Nahla Elsubeihi



Humor is perceived as a mode of discourse and a strategy for human communication. The way humor is constructed in different contexts enriches our understandings of this form of social interaction (Norrick & Chiaro, 2009). It is commonly known amongst humor scholars that the key functions of humor are: constructing personal and group identities, regulating social behavior, and establishing and maintaining relationships (Bell, 2009). However, as we further acknowledge the essential role of humor in social interaction, little is known about how humor fails in conversation. If humor is vital in constructing and sustaining human relationships, its failure may have serious social consequences. A joke that is too threatening to someone’s sense of identity or that is beyond your his/her understanding may elicit failed responses such as anger, incomprehensibility, and non-verbal responses. Therefore, different hearer reaction indicates that the enjoyment of a joke depends to a large extent on the cognitive processes hearers go through when interpreting a joke, their background knowledge, and sociolinguistic variables (i.e. gender, social relationships, and age) (Bell, 2013). As the issue of failed humor is just being explored by pragmaticians, it is significant to explore existent theoretical explanations for the many reasons why humor fails and the ways in which this happens.


The Relationship between Language and Humor

 One of the various ways of provoking laughter is through the use of language. But, in the world of verbal humor, language functions as facilitator. According to Nash (1985), “Through form we come to language, the trigger that detonates the humorous mass” (p.7). Nearly every component of our linguistic system, as a matter of fact, can be exploited for comical needs. Therefore, language structures play a significant role in the creation of ironic discourse (Mahna, 2013).

As the topic of humor has been approached by different scholars, at this point, it is beneficial to give an overview of the research done in the field of verbal humor. In his book, Linguistic Theories of Humor, Attardo (1994) pragmatically defines humor as “a text whose perlocutionary, i.e. intended effect is laughter” (p.7).  In actual fact, it is possible to relate Attardo’s definition to the “speech act theory” proposed by Austin (1962). According to this theory, words not only mean something, but also perform an action. Moreover, utterances perform three acts - one of which is perlocutionary act which is the effect on the hearer. Therefore, in this regard, “a witty story is also an utterance which bears in itself perlocutionary force. The act to which it tries to lead the listener is laughter” (Culler, 1982, p. 113).

Additionally, Bruyn (1993) adds that “jokes are speech acts, acts of humor with perlocutionary effect in mind to amuse” (cited in Mahna, 2012, p. 44). Accordingly, verbal humor can be explored in the framework of Austin’s speech act theory. To illustrate what these humor scholars mean, let us consider the following example:


Clergyman: I now pwonounce you man and wife.

 Bride: And you pwonounce it beautifully, wector (Nash, 1985, p.113).


In this example, both the clergyman and the bride have speech defects. We as hearers may regard it as a joke of mispronunciation. What is significant is that clergyman’s words are a  performative speech act, so the term “pronounce” does not mean “the way it is said,” but rather “what is done with the word.” However, the bride chooses to interpret the word in the former sense (Nash, 1985).


Pragmatics and verbal humor: Inference-oriented approaches

When it comes to the inferential treatments of humor, it is the violation of Grice’s maxims that triggers humor. Grice’s particular concern was how hearers are able to derive the assumptions that speakers implicitly convey. Therefore, in order to infer the implied meaning of an utterance, hearers must engage in some inferential work whose principles lie outside the structure of language (Grice, 1975).  Pragmatic inferences can be made in at least two ways: from the observation of the maxims or from their flouting. In the first case, a speaker may be seen to observe the maxims, yet the hearer will still need to do inferential work. Several authors (e. g. Leech 1983; Yamaguchi 1988) have proposed that humor can arise from the violation of Gricean maxims of conversation. Perchance this was motivated by the observation that cases such as the following remark are often regarded as amusing, Grice himself considered irony -which may be humorous- as a case exploiting the maxim of manner:


Miss X produced a series of sounds that corresponded closely with the score of “Home sweet home” (Grice, 1989, p. 37).


Thus the most obvious implied meaning from this remark, argues Grice, is that Miss X’s performance was weak, and so the speaker may implicate just that.

Irony was not only viewed as a case of violation by Grice - metaphor and hyperbole are cases of violation of the maxim of quality. In situations where it is apparent to both the speakers and their audience that the speakers do not believe what they say, the inference arises that they must be trying to communicate a related proposition - different from the one that is expressed. In the cases of irony, Grice claims that “the most obviously related proposition is the contradictory of the one he purports to be putting forward” (1989, p.34) and so this is derived through implicature.

Attardo (1994) argues that the following cases demonstrate how a humorous effect may rise from the violation of each of the Gricean maxims:

Quantity: “Excuse me, do you know what time it is?” “Yes”

Quality “Why did the Vice President fly to Panama?” “Because the fighting is over”

Relation: “How many surrealists does it take to screw in a light bulb?” “Fish”

Manner “Do you believe in clubs for young people?” “Only when kindness fails” (p.272).  


Attardo (1994) argues that what is violated in these cases is the cooperative principle itself, so that, in Grice’s words, the speaker is seen as “opting out.” Having said that, Attardo poses the following question: “how is it possible then that speakers do successfully engage in communicative practices that involve humorous exchanges?” (p. 275). He goes on to argue that, although these examples are seen non-cooperative, “they still make sense, and are understood and recognized as jokes” (Curco, 1997, p. 133). 

While these aforementioned examples might be funny when uttered in certain situations, it can also be merely rude, or uncooperative – something much different from the “lack of cooperativeness” that leads a hearer to switch to the “non-bona-fide mode of communication” which Attardo takes humor to be. How would a humor recipient, then, know when to interpret an utterance as a sign of rudeness or a humorous remark? These two alternative interpretations are not undesirable for a theory of verbal humor, therefore attempts at humor often fail in precisely this way, and hearers take offence (Curco, 1997).

In order to solve this contradiction, Yamaguchi (1988) proposed the “character-did-it hypothesis”:


(1) One of the characters in the joke is free to violate the maxims of conversation in order to produce the essential ambiguity of the joke. (2) The narrator must avoid violation of the maxims. When for some reason the maxims are to be violated in the narrator’s own report of the event, either the narrator needs to pass on the responsibility for the violation to one of the characters, or at least to minimize the narrator’s own responsibility for the violation in one way or another (p. 327).  

To illustrate this hypothesis, Yamaguchi used this example:

The boss finally agreed to give Ken the afternoon off because he said his girlfriend was going to have a baby. Next morning, the boss said, ‘‘Was it a boy or a girl?’’ ‘‘Too soon to tell’’, replied Ken. ‘‘We won’t know for another nine months.” (p.329).

In this example, Yamaguchi argues that it is Ken and not the narrator that “quietly an unostentatiously” (Grice, 1989, p. 30) violated the Gricean maxim of quantity. Therefore, the humor receivers would assume that the character Ken is responsible for the violation since, according to Yamaguchi, the narrator is only “reporting the violation,” rather than “using” the uncooperative expression (Yamaguchi, 1988).


Relevance Theory

Sperber and Wilson (1981) suggested a revision of some of Grice’s ideas in the early eighties, particularly the maxim-based explanation of figurative language. Therefore, they proposed a reduction of Grice’s conversational maxims to the principle of relevance, which later developed into relevance theory, a theory in contemporary pragmatic research. Regarding the study of humor, there has been several analysts that argued that in order to achieve humorous effects, Gricean maxims, especially the maxim of quantity and relation, must be intentionally violated (Attardo,1990, 1993,1994; Chiaro,1992;Marino,1988;Morreall,1989).

However, the hypothesis of humor as a violation of Grice’s maxims does not fit Sperber-Wilson’s relevance-theoretic approach, as Grice’s view that cooperation is prerequisite is deemed unnecessary. The fact is achieving “optimal relevance” is “less demanding than obeying Gricean maxims” and so it is “possible to be optimally relevant without being ‘as informative as is required’ by the current purposes of the exchange...the degree of co-operation described by Grice is not automatically expected of communicators” (Yus, 2003, p. 1297). As an example, this exchange from the popular American sitcom “The Big Bang Theory,” demonstrates obtaining optimal relevance:


Leonard: Will you please take that stupid hat off.

Howard: No, I want to blend in.

Raj: To what? Toy Story? (Episode 1 in Season 3)


            This particular episode is set in Texas where cowboys are stereotypical images of the State. Apparently Howard wants to blend in, however, by “deliberately dismissing this maximal relevance,” Raj mentions the Toy Story movie. Therefore, in order to obtain optimal relevance, viewers have to further use their encyclopedic knowledge: in the movie Toy Story, Detective Woody also wears a cowboy hat, and so this extra processing effort makes onedeeply feel about the silliness of the hat (Shuqin, 2013).

Given the fact that Grice’s CP does not hold within a relevance-theoretic approach, many researchers of humor (Jodowiec, 1991; Curco, 1997) have rejected the concept of achieving humor through violating Grice’s maxims, and have subscribed to a more “cognitive approach” in which a mental search for an optimally relevant interpretation also “covers the processing of humorous discourse and the derivation of humor effects” (Yus, 2003, p. 1298). Although humorists may be obscure, ambiguous and keep relevant information to themselves to achieve humorous effects, the principle of relevance is invariably applied to both humorous and non-humorous discourse, without having to invoke any maxim or principle of cooperation (Curco, 1997).

In order to be aware that that certain “effort-demanding interpretive paths” are favored by the interlocutor in the eventual humorous effects may be a useful “contextual factor” in the search for an optimally relevant interpretation of humorous discourses (Yus, 2003). Therefore, this realization on the part of the joke receiver, and the joker teller’s control over the number and quality -whether explicit or implicit- of assumptions is an important contextual feature, claims Yus (2003), as it “plays a part in the search for the supposedly intended interpretation of the stimulus and its eventual humorous effects” (p. 1299). To illustrate his argument, he used the following joke:


Q. Why did the cookie cry?

A. Because its mother had been away for so long.

(Pepicello and Green, 1983 cited in Yus, 2003 , p. 1299)


The above canned question-and-answer formula requires several interpretive steps: as the format of the joke alerts the recipient to a forthcoming humorous utterance, “they are warned of the likelihood of irrelevant elements in the joke” (p. 1299)  (for example, the receiver has to assign an animate attribute to a cookie). Next, the lexical features such as “cry” and “mother” make the ‘‘away for so long’’ interpretation most possible; however, the interpretation is not found relevant in the current context since no possible humorous effects can be derived. As a humorous interpretation was intended, this would lead the hearer to further processing, in search for an impetration consistent with the principle of relevance. As a result, a second interpretation is reached:  based on homophony, “away for” would sound like “a wafer.” Having decided that the utterance was intended as a joke, the hearer is then confronted with two interpretations for the same text (a wafer/awayfor), and so “concludes that the two senses are supposed to co-exist humorously” (p. 1299).


Responses to failed humor

 It has been widely acknowledged by humor scholars that humor functions as a means of maintaining human relations, and constructing groups and personal identities. As we recognize the integral role of humor in social interaction, humor failure can have serious social consequences.  Bell (2009) provided a description of the variety of ways in which humor can fail, and how responses to different types of humor failure are constructed. Bell and Attardo (2010) found that humor may fail due to a “hearer’s inability to process the language, understand certain words, understand the pragmatic force of the utterance, recognize the humorous frame, grasp incongruity in a joke, or appreciate the humor” (p.423).  They additionally mention that an incapability to “appropriately support humor” is seen as one type of humor failure. Each of these types of failure may elicit a different type of response. For example, a person who experiences an inability to process the langue due to the existence of noise will surely respond differently from an individual who is able to hear the utterance, but interprets it seriously instead of within the intended playful frame of the speaker (Bell & Attardo, 2010).

 Hay (2001) argues that a good sense of humor is a valued personality trait. Hearers that lack this trait may risk at presenting themselves. This is why, she claims, that a joke that has been found “offensive may be received with both laughter and denials” (p. 58) (e.g. ‘‘That’s not funny.’’). To avoid presenting themselves as mirthless individuals, hearers encountering failed humor must react in a manner that reflects both their “recognition of the utterance as a joke” in addition to their appraisal of it as failed in some respect. Conveying these subtle messages, of course, varies across contexts and amongst interlocutors (Hay, 2001).

In her study of online arguments about rape jokes, Kramer (2011) examined a wider range of responses to these jokes, including the positions of those who defended the jokes as humorous, those who found them comical based on certain conditions, and those who argued that rape jokes can never be amusing. In her study she demonstrated the various “humor ideologies” – the different views about the power of humor and the conditions under which it is fitting to laugh. For example, some of the online respondents used a range of strategies to argue that while “rape itself is not amusing, a joke about rape can be amusing” (Bell, 2013, p. 177).

In her study, Bell (2009) argues that as humor may fail because the hearer may not find it amusing may present a related “interactional dilemma” (p. 146). To demonstrate, she selected a joke that was unlikely to be found amusing and used it to elicit humor failure reactions. She discovered that participants responded in ways that displayed their recognition and understanding of the joke, however was evaluated it “as a poor attempt at humor” (p. 146).  For example, while laughter was the most usual humor response, it was also repeatedly accompanied by metalinguistic comments that disapprovingly judged the joke and/or the joke teller (e.g. ‘‘That is/you are so stupid). Additionally, she discovered that such “aggressive comments” were mainly directed to close intimates, not only as demonstrating their desire to not hear similar jokes in the future, but also “reconstructing their shared identity as one in which such humor is not acceptable” (p.146).

In spite of all this, not all attempts at humor in conversations may receive a response. Priego-Valverde (2009) reveals an uncommon look at failed humor that arose in recordings of casual conversation, some of which failed in the sense that it merely was not acknowledged by respondents. To understand how humor is misconstrued, she draws on Bakhtin’s (1981) model of language and his concept of double-voicing. In her study, she describes that humor is created with two voices: the first voice produces a given utterance, while the second comments on the utterance (possibly mocking the utterance, or ironically responding to it).  It is the second voice, she stresses, that can muddle humorous communication, as the hearer may not be able to distinguish between the voice speakers align themselves with, and therefore to what extent they are joking (Priego-Valverde, 2009). As a participant in the group she recorded, this allowed Priego-Valverde’s to identify two types of failed humor: humor that was not recognized, and humor that was recognized, but ignored. The first type of humor failure is fairly straightforward; the hearers fail to recognize that the speaker intended an utterance to be playful. As for the second type of failure where humor was recognized, but ignored, the hearers align themselves to the serious voice of the joker, and so may be a reaction to failed humor (Priego-Valverde, 2009).

In terms of responses to incomprehensible humor it was Sacks (1974) that addressed this humor failure. His study concentrated on a dirty (sexual) joke told by a teenage boy to two others. As jokes take a narrative form, Sacks claims that all jokes, especially dirty ones, function as ‘‘understanding tests’’ (p. 336), therefore failure to grasp them has different consequences for the hearer than does misunderstanding a serious story. When a serious story is being told, Sacks asserts that misunderstandings perceived by the hearer, a request for clarification can be made with no additional judgment attached to it. On the other hand, if the narrative is a joke, exhibiting a lack of understanding will cause the hearer “to fail the understanding test” and so risk being negatively judged, for example, as ‘‘unsophisticated’’ (Sacks, 1974). Due to the positive value placed on humor as personality trait, Sacks further argues that hearers will be motivated to conceal their inability to make sense of the joke. This concealment can be easily achieved as ‘‘there is available a general way to appropriately respond which can be used whether one understands or not, i.e., laughter produced at the recognized completion’’ (Sacks, 1974, p. 345). Hence, we are led to expect hearers to fake appreciation even though they do not comprehend the utterance. Another response joke recipients adhere to when they experience humor failure, Sacks claims, is placing the blame on the joker, therefore by ridiculing the joke or the teller hides the hearers’ lack of sophistication in understanding the joke (Sacks , 1974).


Sociolinguistic Variables

            Hearer reaction to humor may depend on sociolinguistic variables. For example, the sex of the respondent proved to be a significant variable. In her study, Bell (2013) discovered that men tend toward “more indirect means of expressing their lack of comprehension than women” (p. 184). She argues that this gender-based difference with regard to humor comprehension could be explained by how males and females broadly orient towards humor: “males [tend] to produce and appreciate humor that is more aggressive, where women’s humor tends to be more supportive” (p. 184). In Sacks (1974) study, where his analysis of the telling of a joke was carried among males, if humor was seen as “supportive” than his assertion that jokes function as a knowledge test, there is no need for defensive action, and so direct admission of incomprehensibility can be done with little loss of face (Sacks, 1974).

Moreover, the sex of the joke-teller also plays a significant factor affecting the way people respond. In her analysis, Bell (2013) discovered that the ‘‘repetition of punch line’’ response was frequently directed towards female than at male joke-tellers. In addition, female tellers “received more laughter” and explicit expressions of incomprehensibility than males. Bell also revealed that responses that assessed the joke or the teller of the joke were often received by males, and male tellers were asked to explain the joke more frequently than were females. She concludes these responses may signify that humor recipients tend to perceive a joke told by a male “more as a challenge than that told by a female Hence, the male joke tellers tend to be targeted with (usually negative) assessments” (p.185). Female humorists, on the other hand, were responded to in unchallenging ways, and even received more laughter (though this laughter may not always have been of a supportive quality) (Bell, 2013).

While at a lower level of significance than gender, social relationship also proved to be a noteworthy variable. Bell found out in her analysis that explicit expressions of incomprehensibility and requests for explanations occur more frequently among strangers and intimates. For instance, negative and aggressive comments, according to Bell signaled intimacy. However, vague expressions that “leave some room for interpretation” occurred more frequently between acquaintances. These expressions entailed laughter, interjections (e.g. ‘‘Wait, what?!’’), and questions about the utterance (Bell, 2013).

In terms of the variable of age, Bell’s (2013) study found that interjections as a response to the joke were used much more frequently by a younger group than by an older group. Additionally, she revealed that the former tend to laugh, and assess the joke and its teller more frequently. The older group, however, “favored the direct strategies of explicitly stating their problems in understanding and of requesting explanations” (p. 186). Barbieri’s (2008) corpus-based study also examined age-based variation in American English. In comparing the marked difference between speakers aged 15--25 to that of speakers aged 35—60, she claims that “the instantiation of speaker’s stance, emotional involvement, and intersubjectivity appears to play a much bigger role among the youth, than among ‘mature’ adults, suggesting that adults are relatively less inclined to overt displays of feelings, attitudes, and emotions’’ (p. 79). Barbieri’s analysis of linguistic elements also overlapped with Bell’s (2013) with respect to interjections, which she discovered to be more common amongst younger groups, therefore their overt emotionality may be intended to signal strong involvement (Bell, 2013).



In order to gain a deeper understanding of verbal humor, pragmatics and cognitive analyzing methods help us to better appreciate this linguistic phenomenon. Sperber and Wilson’s RT is pragmatic theory not particularly designed for the analysis of humor, however its theoretical hypotheses best explains how humorous interpretations are made, and is therefore applicable to the interpretation of all types of discourse, including humorous discourses (Curco, 1997; Yus, 2003). Although the pragmatics study of humor, how we understand it, and how it succeeds has been extensively researched, research on humor failure has been continued to be neglected (Bell, 2009, 2013). Therefore, in order to build and maintain human interactions, research on this phenomenon should be increasingly encouraged.



Attardo, S. (1990). The violation of Grice’s maxims in jokes. In K. Hall,. et al., (Eds.), Proceedings of the 16th Berkeley linguistics society conference, (pp. 355–362).

Attardo, S. (1993) Violation of conversational maxims and cooperation: The case of jokes. Journal of Pragmatics. 19, 537–558.

Attardo, S. (1994). Linguistic theories of humor. Berlin: Mouton de Gruyter.

Bakhtin, M. (1981). The dialogic imagination. Austin :University of Texas Press.

Bell, N. D. (2009). Responses to failed humor. Journal of Pragmatics. 41(9), 1825-1836. doi:10.1016/j.pragma.2008.10.010

Bell, N. D., Attardo, S. (2010). Failed humor: Issues in non-native speakers’ appreciation and understanding of humor. Intercultural Pragmatics. 7(3), 423-447.

Bell, N. D. (2013). Responses to incomprehensible humor. Journal of Pragmatics. 57, 176-189. doi:10.1016/j.pragma.2013.08.019

Barbieri,F.(2008) Patterns of age-based linguistic variation in American English. Journal of Sociolinguistics. 12(1), 58–88.

Chiaro, D. (1992). The language of jokes. Analysing verbal play. London: Routledge.

Curco. C. (1997). The pragmatics of humorous interpretations: a relevance-theoretic approach. Doctoral thesis, University of London.

Grice, P. (1975). Logic and conversation. In P. Cole, & J. Morgan (Eds.), Syntax and semantics, (pp. 41–58). London: Academic Press.  

Girce, H. P. (1989). Studies in the way of words. Cambridge, Mass: Harvard University Press

Hay, J. (2001). The pragmatics of humor support. Humor: International Journal of Humor Research, 14 (1), 55-82.

Kramer, E. (2011). The playful is political: The metapragmatics of internet rape-joke arguments. Language in Society 40, 137-168.

Jodowiec, M. (1991). The role of relevance in the interpretation of verbal jokes: A pragmatic analysis. Unpublished PhD Dissertation. Krakow: Jagiellonian University,

Marino, M. (1988). Puns: the good, the bad, and the beautiful. Humor, 1(1), 39-48.

Morreall, J. (1989). Enjoying incongruity. Humor, 2(1), 1–18.

Nash, W. (1985) The language of humour, style and technique in comic discourse. London and New York: Longman

Norrick, N. R., & Chiaro, D. (2009). Humor in interaction. Amsterdam; Philadelphia: John Benjamins Pub. Co.

Priego-Valverde, B. (2009). Failed humor in conversation: a double-voicing analysis. In D. Chiaro, & N. Norrick (Eds.), Humor in interaction (pp. 165-183). Amsterdam: John Benjamins.

Sacks, H. (1974). An analysis of the course of a joke’s telling. In B. Richard, J. Sherzer, (Eds.), Explorations in the ethnography of speaking (pp. 325-345). Cambridge: Cambridge University Press, Cambridge.

Shuqin, H. (2013). A relevance theoretic analysis of verbal humor in the big bang theory. Studies in Literature and Language, 7(1), 10.

Sperber, D. and Wilson, D. (1981). Irony and the use-mention distinction. In P. Cole (Ed.), Radical pragmatics. New York: Academic Press

Yus, F. (2003). Humor and the search for relevance. Journal of Pragmatics, 35(9), 1295-1331. doi:10.1016/S0378-2166(02)00179-0



Nahla Elsubeihi is a Libyan student at AUS. She majors in English Language and is in her final year. Upon graduating, she plans to amplify her knowledge and expand her career options by pursuing a higher degree in Linguistics. She is specifically interested in sociolinguistics and discourse analysis. She is committed to pursue a career in the world of academia and pedagogy. Nahla is a voracious reader, an intersectional feminist, and her favorite movie catchphrase is “Leave the gun. Take the cannoli.”

bottom of page