Towards a neural circuit model of verbal humor processing: An fMRI study of the neural substrates of incongruity detection and resolution
Highlights
► This study sought to isolate the neural substrates underlying humor comprehension. ► Incongruity detection and resolution were segregated using a ‘nonsensical’ condition. ► Detection activated the right middle temporal gyrus and right medial frontal gyrus. ► Resolution activated the left superior frontal gyrus and left inferior parietal lobe. ► A neural circuit model of humor comprehension and elaboration processing is proposed.
Introduction
Humor is unique to mankind and plays an important role in social settings. However, the neurological mechanisms underlying humor comprehension are still not fully understood. In recent years, new technologies have made it possible to develop progressively more refined understandings of these mechanisms. The present study seeks to further contribute to their identification through a novel combination of functional magnetic resonance imaging (fMRI) technology, experimental design, and verbal stimuli constructed to distinguish between the key stages through which the brain processes humor.
The present study builds on earlier research making use of the framework provided by Wyer and Collin's comprehension–elaboration theory of humor (Chan et al., 2012, Wyer and Collins, 1992). The comprehension–elaboration theory of humor claims that humor processing can be segregated into two phases, comprehension and elaboration. Comprehension includes both the experience of surprise resulting from encountering unexpected or incongruous information and the re-establishment of coherence which results when the unexpected information is reinterpreted using concepts and schemata from a different knowledge domain. The elaboration phase involves cognitive elaboration of the implications of the reinterpreted situation and subsequent inducement of the feeling of amusement.
Our previous study attempted to differentiate the respective brain areas subserving these two phases in the processing of verbal jokes, by comparing three conditions: funny, unfunny, and garden path. In that study, the bilateral inferior frontal gyri and left superior frontal gyrus were found to be associated with humor comprehension, while the cortical region of the left ventromedial prefrontal cortex and the subcortical regions in the bilateral amygdalae and bilateral parahippocampal gyri were found to be responsible for the feeling of amusement during the elaboration process (Chan et al., 2012). The present study seeks to further advance our understanding of the comprehension of verbal humor, by distinguishing the neural substrates of incongruity detection and incongruity resolution. It is expected that our findings will advance our understanding of the neurological mechanisms underlying humor processing and further specify a neural circuit model of verbal humor processing involving three stages: incongruity detection, incongruity resolution, and elaboration.
While a great deal of the humor that we experience on a daily basis is verbal in nature, many studies using fMRI to study humor processing have focused on non-verbal processing, for example, using nonverbal visual cartoons (Azim et al., 2005, Bartolo et al., 2006, Moran et al., 2004, Samson et al., 2008, Samson et al., 2009, Wild et al., 2006) or humorous videos (Moran et al., 2004, Neely et al., 2012). This study joins earlier studies (Bekinschtein et al., 2011, Goel and Dolan, 2001, Ozawa et al., 2000, Uekermann et al., 2006, Watson et al., 2007) in focusing on verbal humor processing.
Within the comprehension–elaboration framework, comprehension can be further divided into two stages, incongruity detection and incongruity resolution (Suls, 1972). For humor to be comprehended as such, it must first be surprising in some way; that is, some incongruity with the preceding context must be detected. Then, coherence must be restored or, in other words, the incongruity must be resolved. This can be illustrated through consideration of the traditional structure of a joke, which includes a setup and a punch line. The setup is a short statement providing the details necessary to follow the joke, establishing a context which enables expectations to be formed concerning what is likely to follow. The punch line then generates humor by introducing an unexpected ‘twist’ which in some way violates the expectations established by the setup (Attardo, 1997, Vaid et al., 2003, Wild et al., 2006). In processing the joke, the reader thus must first respond to the setup by establishing a set of expectations, then upon reading the punch line must detect an incongruity between the punch line and the setup, and then resolve this incongruity in a playful manner.
A growing number of studies have made use of fMRI technology to isolate and identify different neural regions involved in humor processing and a progressively more precise and detailed account of the brain mechanisms underlying humor appreciation has emerged (e.g., Bartolo et al., 2006, Bekinschtein et al., 2011, Goel and Dolan, 2001, Moran et al., 2004, Neely et al., 2012, Wild et al., 2006). As noted, most previous fMRI studies of humor have focused on non-verbal humor, with only a relatively small number focusing on the processing of verbal humor (Bekinschtein et al., 2011, Goel and Dolan, 2001, Ozawa et al., 2000, Uekermann et al., 2006, Watson et al., 2007). One suggestion to emerge from these studies is that the left inferior parietal lobule (IPL) may be active in the semantic processing of verbal humor (Ozawa et al., 2000).
Most previous fMRI studies of humor have used a funny condition and an unfunny (non-funny) condition as stimuli (Azim et al., 2005, Bartolo et al., 2006, Bekinschtein et al., 2011, Goel and Dolan, 2001, Mobbs et al., 2003, Watson et al., 2007, Wild et al., 2006). It appears that the inferior frontal gyrus (Azim et al., 2005, Bartolo et al., 2006, Bekinschtein et al., 2011, Goel and Dolan, 2001, Mobbs et al., 2003, Watson et al., 2007), middle temporal gyrus (Bartolo et al., 2006, Moran et al., 2004), and superior frontal gyrus (Bekinschtein et al., 2011) are associated with the humor comprehension process. However, designs limited to comparisons of funny and unfunny conditions have not been able to fully differentiate the neural substrates involved in the incongruity detection and resolution stages of comprehension processing.
The present attempt locates itself within a series of studies which have sought to distinguish between the different aspects of comprehension processing. Brownell et al. (1983) administered a joke-completion task to 12 patients with right-hemisphere damage (RHD) and 12 matched normal controls. Their participants read the setups to a series of 16 jokes and were asked to choose the best punch lines from a set of four alternatives, which included the funny ending (requiring incongruity detection and resolution), a nonsequitur ending (incongruity detection without resolution), a straightforward neutral ending and a straightforward sad ending (no incongruities to detect and resolve). RHD patients were less likely to choose the correct funny ending and more likely to endorse the nonsequitur endings. The present study is similar in that it also includes funny, nonsensical (‘nonsequitur’) and unfunny (‘straightforward’) conditions; however, it differs in that all participants were neurologically normal volunteers instead of patients.
In Bekinschtein et al. (2011)'s fMRI study, 18 participants listened to 23 each of ambiguous jokes (puns), ambiguous sentences (containing a key word with dual meanings), unambiguous jokes and unambiguous sentences. The processing of stimuli containing ambiguous words (presumably involving the detection and resolution of an incongruity for at least one of the meanings of the words) activated the left interior temporal gyrus (ITG); for the ambiguous jokes, the left IFG, right upper MTG, and left SFG were also activated in a whole-brain analysis. The interaction between ambiguity and jokes revealed a significant activation in the left anterior IFG, spreading to the middle frontal gyrus (MFG). As with their study, the present study seeks to refine our understanding of the how ambiguities are processed using verbal stimuli. It contains an additional ‘nonsensical’ condition, introduced in order to differentiate incongruity detection from incongruity resolution.
Samson et al. (2008) used funny cartoons and nonsense cartoons (cartoons containing an irresolvable incongruity). They found activation in the bilateral supramarginal gyri during successful incongruity resolution, with activity in the right middle frontal gyrus (rMFG) in instances where an incongruity was detected but not resolved. The present study differs from these and other similar attempts (Marinkovic et al., 2011, Samson et al., 2009) to identify the neural bases of incongruity detection and resolution during humor comprehension by making pairwise comparisons using carefully matched funny, unfunny, and nonsensical verbal stimuli. Stimuli in the nonsensical condition contain incongruities which are irresolvable and carefully matched with the other two conditions, in order to better isolate the distinct substrates associated with incongruity detection and resolution.
To summarize, the present study seeks to further specify a three-stage neural circuit model of verbal humor processing, which includes incongruity detection, incongruity resolution and elaboration. It employs an event-related fMRI experiment to identify the neural substrates associated with the detection and resolution of incongruities during the comprehension of the humor in verbal jokes by comparing three carefully paired conditions: unfunny (non-funny or congruous; the baseline condition), nonsensical (irresolvably incongruous) and funny (resolvably incongruous; i.e., humorous). For the verbal stimuli in the unfunny condition, there is no incongruity between the setup and punch line that needs to be detected and resolved. Therefore, no surprise or amusement is expected in this condition. In the nonsensical condition, the punch lines are incongruous and puzzling, but there is no resolution, as there is no logical connection between the setups and punch lines (Shultz, 1974). Compared to the unfunny condition, the nonsensical condition contains an additional, incongruous element and should thus elicit the cognitive operations needed to detect this incongruity. Based on earlier findings of involvement of the right medial temporal gyrus in the detection of semantic violations (Kuperberg et al., 2000, Newman et al., 2001, Ni et al., 2000), and of the right MFG in humor processing (Azim et al., 2005, Samson et al., 2008) and context monitoring (Hampshire et al., 2009, Petrides, 2005), we predict that these regions will be involved in the incongruity detection process.
Finally, the funny condition contains a resolvable incongruity. In addition to the operations required to detect the incongruity, it should also call forth those required to successfully resolve the incongruity and, thereby, to comprehend the humor (Fig. 1). Our earlier study (Chan et al., 2012) found activation of the bilateral IFG during humor comprehension; other research has found greater activation in the left IFG (Bekinschtein et al., 2011, Rodd et al., 2005 Samson et al., 2008) and right IFG (Moran et al., 2004) related to the resolution process. Similarly, our earlier study found activation of the left SFG during humor comprehension, and other research on humor processing has associated this area with incongruity resolution (Bekinschtein et al., 2011). Finally, the left inferior parietal lobe has been associated with semantic integration and coherence (Chou et al., 2006, Chou et al., 2009). We therefore predict that these regions (bilateral IFG, left SFG and left IPL) will be associated with incongruity resolution. This successful resolution completes the comprehension stage and leads on to the elaboration stage (Fig. 2). In the present study, the setups were the same across all three conditions, with only the punch lines varying, so that the observations should reflect the new information in the punch line with the schemata pre-activated by the setup.
The present study thus made use of a different design in an attempt to differentiate the respective brain areas corresponding to humor comprehension in the processing of verbal jokes, by comparing three conditions: funny, unfunny, and nonsensical. Based on the comprehension–elaboration framework (Wyer and Collins, 1992), incongruity-resolution theory (Suls, 1972), and the finding of previous neuroimaging studies of humor processing, we hypothesized that (non-humorous) nonsensical stimuli, in comparison with the unfunny stimuli, would elicit increased activation related to incongruity detection in several brain regions, including the right middle temporal gyrus and right medial frontal gyrus. We also predicted that the funny stimuli, in comparison with the nonsensical stimuli, would elicit increased activation related to resolution of incongruity in several brain regions, including the bilateral inferior frontal gyrus, left superior frontal gyrus and left inferior parietal lobule.
Section snippets
Participants
Twenty two neurologically healthy volunteers [9 males and 13 females aged 20–29 years, 21.86 ± 2.17 years (mean ± SD)] participated in this study. They were all right-handed (as determined by the Edinburgh Handedness Inventory; Oldfield, 1971), native Mandarin speakers. The study was approved by the Research Ethics Committee of the National Taiwan University Hospital, and all of the subjects gave their informed consent to participate before commencing the experiment.
Stimuli
Sixty-four stories in Mandarin
Behavioral results
In terms of comprehensibility ratings during the scanning, the mean rating scores for comprehensibility were 93.71% for the funny condition, 89.31% for the unfunny condition, and 8.29% for the nonsensical condition. A one-way repeated-measures ANOVA on participants' rating scores was significant, F (2, 42) = 948.27, p < .001. Bonferroni post-hoc tests revealed that the funny and unfunny conditions were more comprehensible than the nonsensical condition. There was no significant difference in the
Discussion
In the present study, an event-related fMRI experiment was used to identify and distinguish the neural substrates of the incongruity detection and resolution processes during comprehension of verbal jokes. The activations believed to be related to the detection of incongruities were identified by comparing a nonsensical condition with a paired unfunny condition. The results confirm that the right MTG and the right MFG are involved in processing irresolvable incongruities, where there is no
Acknowledgments
The work was supported by the Ministry of Education, Taiwan, under the Aiming for the Top University Plan at National Taiwan Normal University. We also thank the National Science Council for funding this study, through projects on standard stimuli and normative emotional responses in Taiwan (NSC-97-2420-H-002-220-MY3), neuropsychological traits of mathematically and scientifically talented students with and without Asperger's syndrome and the training effects of sense of humor on them (NSC
References (50)
- et al.
Comprehension of humorous and nonhumorous materials by left and right brain-damaged patients
Brain Cogn.
(1986) - et al.
Surprise but not coherence: sensitivity to verbal humor in right-hemisphere patients
Brain Lang.
(1983) - et al.
Knowledge of object manipulation and object function: dissociations in apraxic and nonapraxic subjects
Brain Lang.
(2002) - et al.
Segregating the comprehension and elaboration processing of verbal jokes: an fMRI study
NeuroImage
(2012) - et al.
Representaion of manipulable man-made objects in the dorsal stream
NeuroImage
(2000) - et al.
Frequency of concrete words modulates prefrontal activation during semantic judgements
NeuroImage
(2002) - et al.
Developmental changes in the neural correlates of semantic processing
NeuroImage
(2006) - et al.
Time course of semantic processes during sentence comprehension: an fMRI study
NeuroImage
(2007) Bilateral brain processes for comprehending natural language
Trends Cogn. Sci.
(2005)- et al.
Develpmental changes in the inferior frontal cortex for selecting semantic representations
Dev. Cogn. Neurosci.
(2011)
Lexical ambiguity in sentence comprehension
Brain Res.
Humor modulates the mesolimbic reward centers
Neuron
Neural correlates of humor detection and appreciation
NeuroImage
The assessment and analysis of handedness: the Edinburgh inventory
Neuropsychologia
Guidelines for reporting an fMRI study
NeuroImage
Neural substrates of incongruity-resolution and nonsense humor
Neuropsychologia
A two-stage model for the appreciation of jokes and cartoons: an information-processing analysis
Getting a joke: the time course of meaning activation in verbal humor
J. Pragmat.
The semantic foundations of cognitive theories of humor
Humor: Int. J. Humor Res.
Sex differences in brain activation elicited by humor
Proc. Natl. Acad. Sci.
Humor comprehension and appreciation: an fMRI study
J. Cogn. Neurosci.
Why clowns taste funny: the relationship between humor and semantic ambiguity
J. Neurosci.
Neural correlates of sensory and decision processes in auditory object identification
Nat. Neurosci.
Functions of the left superior frontal gyrus in humans: a lesion study
Brain
The role of inferior frotnal gyrus and inferior parietal lobule in semantic processing of Chinese characters
Exp. Brain Res.
Cited by (107)
The differential processing of verbal jokes by neural substrates in indigenous and Han Chinese populations: An fMRI study
2024, Behavioural Brain ResearchHypo- and hyper-activation in frontotemporal lobe during humor processing in patients with first episode schizophrenia
2024, Asian Journal of PsychiatryMoney or funny: Effective connectivity during service recovery with a DCM-PEB approach
2023, Biological PsychologyDifferential neural substrates for responding to monetary, sexual humor, and erotic rewards
2022, Biological PsychologyDynamic impact of intelligence on verbal-humor processing: Evidence from ERPs and EROs
2022, Journal of Neurolinguistics