Dynamic impact of intelligence on verbal-humor processing: Evidence from ERPs and EROs
Introduction
Humor, involving complicated high-level cognitive activities, is unique to humans. Due to the positive effects of humor on both physical health and mental health, researchers have conducted a number of studies by relating humor with a wide range of influencing factors, ranging from external factors relevant to society, family, school and workplace, to internal factors, such as gender (Chang, Ku, & Chen, 2018), age (Greengross, 2013), personality (Martin, 2003; Samson, Hempelmann, Huber, & Zysset, 2009), emotional quotient (EQ) (Gignac, Karatamoglou, Wee, & Palacios, 2014) and intelligence quotient (IQ) (Vrticka, Black, Neely, Shelly, & Reiss, 2013; Christensen, Silvia, Nusbaum, & Beaty, 2016, 2018). The synergetic effects of these factors lead to the appreciation of humor, suggesting the complex nature of humor processing. The present study will mainly consider the role of intelligence (as measured by IQ) to reveal the correlates of general intelligence with only one kind of humor material, verbal humor, through Event-related Potentials (ERPs) and Event-related Oscillations (EROs).
Based on the incongruity-resolution model of Suls (1972), to get a joke needs three stages: first, the receiver detects the contradictions between the information and his or her expectations; then he/she has to find a novel way to make the information consistent, finally, resulting in the feeling of mirth. By means of electroencephalogram (EEG) techniques, verbal-humor processing was further explored and dissociated into three stages: incongruity detection, incongruity resolution and amusement/mirth (Chan, Chou, Chen, & Liang, 2012; Chan et al., 2013; Feng, Chan, & Che n, 2014; Shibata, Terasawa, & Umeda, 2014; Du et al., 2013; Chang et al., 2018). However, due to some task-dependent reasons caused by characteristics of the stimuli employed in different experiments, there are some inconsistencies among the experimental results on the neural mechanism of verbal-humor processing in different stages.
ERPs, a physiological measure of mass neuronal activity, has been frequently used to examine the functional organization of the brain for language processing (Kutas & Van Petten, 2006). Across studies, some ERP components are related to the three stages of verbal-humor processing including N400/Left-Anterior Negativity (LAN) (corresponding to incongruity-detection stage), sustained LAN/P600 (corresponding to incongruity-resolution stage) and Late Positive Component (LPC) (corresponding to mirth stage).
For the first stage, the incongruity-detection stage, the humorous materials elicit greater amplitude than non-humorous materials at about 400 ms after stimulus onset, which is depicted as the N400 effect. N400 itself is just a more negative-going deflection than that seen to other conditions, and its distribution is maximal over centro-parietal electrode sites, although the distribution can change slightly depending on the eliciting stimulus (Kutas & Federmeier, 2011). In language processing, N400 has been found to be modulated by how much a word matches contextual expectations (Van Petten & Luka, 2012). In verbal-humor processing, N400 is often deemed to reflect the incongruity detection, relevant to the violations of predictability and semantic congruence (Chang et al., 2018; Coulson & Kutas, 2001; Du et al., 2013; Feng et al., 2014; Ku, Feng, Chan, Wu, & Chen, 2017). Usually, the prototypic N400 semantic incongruity effect is broadly distributed across the scalp, larger over parietal, posterior temporal, and occipital sites than frontal sites (Kutas and Van Petten, 2006). But some verbal-humor related ERP studies found that similar negative deflections at about 500 ms were activated by left-anterior electrodes, and thus, researchers depicted them as Left-Anterior Negativity component (LAN) (Canal et al., 2019; Coulson & Kutas, 2001; Coulson & Lovett, 2004; Coulson & Williams, 2005), which was to answer difficult-to-process or flat-out ungrammatical questions when a working memory system becomes overburdened from trying to maintain both lexical and filler-gap information simultaneously (Kutas and Van Petten, 2006). These findings are closely related to the incongruity detection in the first stage of verbal-humor processing.
For the second stage, the incongruity-resolution stage, compared with non-humorous materials, humorous materials often activate greater positivity after 500 ms in multiple areas, which is considered the P600 component. P600s are language-relevant ERPs and are thought to be elicited by grammatical errors and other syntactic anomalies, appearing mostly on centro-parietal electrodes, but frontal P600s have also been observed (Kaan & Swaab, 2003). In verbal-humor processing, P600 reflects the stage of incongruity resolution between setups and punchlines to reorganize existing information into a new schema retrieved from the long-term memory (Coulson & Kutas, 2001; Coulson & Lovett, 2004; Du et al., 2013; Feng et al., 2014; Ku et al., 2017; Shibata et al., 2017). But some experiments found greater negativity in this stage in the same time window over anterior left lateral sites for verbal-humor stimuli for at least some subgroups of the experimental sample (Coulson & Kutas, 2001; Coulson & Lovett, 2004; Coulson & Williams, 2005), which was deemed sustained LAN (Canal et al., 2019) reflecting the frame-shifting needed to re-establish the coherence (Coulson & Kutas, 2001).
For the third stage, the mirth or amusement stage, humorous stimuli, compared with non-humorous ones, elicited more positive potentials during the time window of about 800–1500 ms, which is regarded as the Late Positive Potentials (LPP). LPP is a sustained positivity reflecting further elaboration and mirth experiences, as it can be modulated by emotional relevance (Liu, Huang, McGinnis-Deweese, Keil, & Ding, 2012; Sabatinelli, Keil, Frank, & Lang, 2013) and positively correlated with emotional arousal (Gierych, Milner, & Michalski, 2005). In verbal-humor processing, LPP has been indexed by a centro-parietal late positive potential (Feng et al., 2014; Ku et al., 2017), referring to the third stage, mirth, caused by emotional processing (Chan et al., 2012; Du et al., 2013; Feng et al., 2014; Gierych et al., 2005; Tu et al., 2014) and specifically generated in the ventromedial prefrontal cortex, bilateral amygdala, and bilateral parahippocampal gyri (Chan et al., 2012). But some other humor-related experiments did not find similar ERP components in the same time window.
EROs are obtained by “decomposing EEG signals into phase and magnitude information represented by frequencies and then changes of frequencies are computed by their energy/power and phase relationships over a millisecond time scale” (Ehlers, Kaneko, Pobledo, & Lopez, 1994), consisting in averaging together a number of EEG epochs, time-locked to the onset of the critical events (Luck, 2014). EROs are also related to stimulus processing (Bastiaansen, Mazaheri, & Jensen, 2011), providing comprehensive framework where oscillatory changes in different frequency bands are linked to specific functional mechanism in language processing (Meyer, 2018; Weiss & Mueller, 2012). The investigations into EROs are becoming increasingly important in neuroscience in recent years, since oscillatory activities in various frequency bands may reflect different aspects of cognition processing (Karakaş & Başar, 2006). Specifically, event-related power changes in theta frequency range (3–7 Hz) showed a marked reactivity to various linguistic manipulations, suggesting a functional role during language processing with its power increases being observed over left occipital areas, bilaterally over the temporal cortex and over frontal-central regions (Hald, Bastiaansen, & Hagoort, 2006). Additionally, a recent review about predictions in language comprehension indicated the increased power in beta range (12–25 Hz) of EEG in the discourse representation (Weiss & Mueller, 2012), while a drop in beta power was seen when encountering incongruent information (Segaert, Mazaheri, & Hagoort, 2018). However, the studies of EROs related with verbal-humor processing are rare and only one recent study which can be referenced indicted that the changes in beta-band power may be associated with the last stage of processing of the joke between 700 and 900 ms (Canal et al., 2019).
People good at telling jokes are normally believed as clever, so humor is viewed as an indicator of wit and hence intelligence (Greengross & Miller, 2011; Kaufman, Kozbelt, Bromley, & Miller, 2008), which has been supported by some empirical studies finding positive correlations between verbal-humor production and intelligence levels (Christensen et al., 2016; Greengross & Miller, 2011; Howrigan & MacDonald, 2008; Kellner & Benedek, 2017). To our knowledge, very few neural studies regarding differences in verbal-humor processing among people with different intelligence levels were in evidence.
There has been continuous interest in the relationship between brain activities and intelligence over the last decade. Growing attention has been paid to the early potentials, especially for the P200 component. For example, a number of studies found shorter latencies (Burns, Nettelbeck, & Cooper, 2000) and greater amplitude of P200 were associated with higher intelligence (Zhang, Caryl, & Deary, 1989). In addition, ERO studies found that subjects with higher intelligence generally exhibited significantly stronger activations in low-frequency oscillations (e.g., the theta band) in cognition tasks (Doppelmayr et al., 2005). The EEG source location for high intelligence group was more anterior, whereas for the average intelligence group the source location was more posterior in theta frequencies (Klimesch, 1996).
Intelligence has previously been shown positively related to the capacity for incongruity detection and resolution during humor processing in children as well as adults by behavioral experiments (Mahony & Mann, 1992; Wierzbicki & Young, 1978). However, such a conclusion lacks neural evidence to strengthen its validity. The only neural evidence of the impact of intelligence on humor processing was targeted at children, implying that higher intelligence may not only facilitate incongruity detection and resolution, but also the subsequent feeling of mirth (Vrticka et al., 2013). But the correlations between verbal-humor processing and intelligence levels in adults is still unclear, and our knowledge of the neural oscillatory processes of verbal-humor is limited. In the current study, we employed the high temporal resolution of the EEG technique to examine the stage at which intelligence differences in neural processing of verbal-humor arise.
The current study aims to use the EEG technique to explore whether the intelligence level has effects on verbal-humor processing, and if it does, our study is to contribute further to empirical evidence about how verbal-humor processing occurs in the brains of people with intelligence differences by examining real-time neural dynamics in light of the three-stage neural circuit model. We developed our hypothesis on the basis of the generally reported intelligence difference observed during humor-processing.
To be specific, firstly, we expected to find significant differences among experimental conditions in three stages of verbal-humor processing related to N400/LAN, P600/sustained LAN and LPP for both higher-level intelligence group and lower-level intelligence group; secondly, we expected to find the earliest significant differences between groups in about 200 ms relevant to the P200 component because of close correlations between latencies/amplitudes of P200 and intelligence levels (Burns et al., 2000; Ma & Li, 2007; Zhang et al., 1989); thirdly, we expected to find group differences in processing pattern in each stage in terms of ERPs, examining people with intelligence differences to follow a unitary-pattern or a dual-pattern model in processing verbal humor; finally, we expected to find more evidence on oscillatory changes underlying the neural mechanism in theta band or beta band and the role of the intelligence level in verbal-humor processing from the perspective of ERO based on relevant previous studies (Canal et al., 2019; Hald et al., 2006).
Section snippets
Participants
A total of 28 college students (15 F) aged between 18 and 24 years (mean age, 21.91 years) were recruited as subjects, who were native speakers of Mandarin Chinese and had no history of current or past neurological or psychiatric illness. All of them were right-handed and had normal or corrected-to-normal vision. Before the experiment, all the subjects completed the written informed consent. The experiment was approved by the Research Ethics Committee of Dalian University of Technology. Two
Behavioral results
The data analysis for groups (RHI/RLI) × conditions (funny/unfunny/unrelated) with a two-way repeated measure ANOVA by SPSS 19 were performed on Response Times (RT) and accuracy respectively, and t tests were used for post hoc analysis. The summary of RT was illustrated in Table 2. The main effect of conditions was significant (F (2, 48) = 6.374, p = 0.005, = 0.210), which reconfirmed the discriminative validity among conditions used in this study, but no group effect nor interaction effect
Discussion
The current study tried to elucidate the dynamic impact of intelligence on different stages of verbal-humor processing from the perspective of ERPs and EROs. Verbal-humor processing is a complex phenomenon concerned with both external factors and internal factors on the basis of extensive language processing, so implicit IQ differences may be minute and difficult to reveal. In the behavioral domain, there was no significant difference between RHI group and RLI group in RT to stimulus
Conclusion
We sought to explore the dynamic impact of IQ on three stages of verbal-humor processing from ERP and ERO perspectives in the current study. The findings showed that the intelligence had dynamic impact on the processing patterns in two stages: incongruity detection and mirth. At the global level, the results indicated that people with different intelligence levels may employ dual-pattern model rather than unitary-pattern model in dealing with these two stages. Namely, in the stage of
Author statement
WH and SJ conceived of the presented idea. LX and WH developed the theoretical formalism. SJ and WH constructed the materials and collected the data. LX, XQ, ZG and WX analyzed the data. ZG and WX technically supported data collection. LX, WH, XQ and SJ interpreted the data. LX wrote the manuscript with support from WH and XQ and coordinate the writing of subsequent versions. All authors contributed to this paper both substantially and formally, and approved the final version of the manuscript.
Acknowledgments
This work was supported by the National Social Science Foundation of China (Grant No. 21BYY102).
References (86)
- et al.
Human anterior and frontal midline theta and lower alpha reflect emotionally positive state and internalized attention: High-resolution EEG investigation of meditation
Neuroscience Letters
(2001) - et al.
EEG correlates (event-related desynchronization) of emotional face elaboration: A temporal analysis
Neuroscience Letters
(2006) - et al.
Event-induced theta responses as a window on the dynamics of memory
Cortex
(2003) - et al.
Human brain EEG indices of emotions: Delineating responses to affective vocalizations by measuring frontal theta event-related synchronization
Neuroscience & Biobehavioral Reviews
(2011) - et al.
Event-related potential correlates of some human cognitive ability constructs
Personality and Individual Differences
(2000) - et al.
‘Honey, shall I change the baby?–well done, choose another one’: ERP and time-frequency correlates of humor processing
Brain and Cognition
(2019) - et al.
Segregating the comprehension and elaboration processing of verbal jokes: An fMRI study
NeuroImage
(2012) - et al.
Towards a neural circuit model of verbal humor processing: An fMRI study of the neural substrates of incongruity detection and resolution
NeuroImage
(2013) - et al.
Sex differences in humor processing: An event-related potential study
Brain and Cognition
(2018) - et al.
Tensor decomposition of EEG signals: A brief review
Journal of Neuroscience Methods
(2015)