Humor processing, mentalizing, and executive function in normal aging

JENNIFER UEKERMANN; SHELLEY CHANNON; IRENE DAUM

doi:10.1017/S1355617706060280

Humor processing, mentalizing, and executive function in normal aging

Published online by Cambridge University Press: 22 March 2006

JENNIFER UEKERMANN ,

SHELLEY CHANNON and

IRENE DAUM

Show author details

JENNIFER UEKERMANN: Affiliation:
Department of Psychology, Institute of Cognitive Neuroscience, Ruhr-University of Bochum, Bochum, Germany
SHELLEY CHANNON: Affiliation:
Department of Psychology, University College London, London, United Kingdom
IRENE DAUM: Affiliation:
Department of Psychology, Institute of Cognitive Neuroscience, Ruhr-University of Bochum, Bochum, Germany

Article contents

Abstract
INTRODUCTION
METHOD
RESULTS
DISCUSSION
References

Rights & Permissions

Abstract

Recent investigations have emphasized the importance of the prefrontal cortex for humor processing. Although the prefrontal cortex is thought to be affected by normal aging, relatively little work has been carried out to investigate the effects of aging on humor processing. In the present investigation participants in three age groups were assessed on a humor comprehension task. They then answered mentalistic and nonmentalistic questions. Executive tasks were also administered. The older group selected significantly fewer correct punchlines from alternatives than the other groups. They were also poorer at answering mentalistic questions, but did not differ significantly for nonmentalistic questions. The findings of the present investigation showed altered in humor processing in normal aging, and this appeared to be related to mentalizing ability. (JINS, 2006, 12, 184–191.)

Keywords

Theory of mind Mentalizing Cognitive Development Social cognition Incongruity Resolution

Type: Research Article
Information: Journal of the International Neuropsychological Society , Volume 12 , Issue 2 , March 2006 , pp. 184 - 191

DOI: https://doi.org/10.1017/S1355617706060280 [Opens in a new window]
Copyright: © 2006 The International Neuropsychological Society

INTRODUCTION

Despite its important role in human social interaction, only a few neuropsychological investigations on humor processing exist. Studies conducted so far have been mainly based on the incongruity resolution theory (Suls, 1972). Incongruity resolution theory proposes two stages of humor processing. The first stage (incongruity detection) refers to the perception of an incongruous element, which is resolved in the second stage (resolution).

Early studies have compared patients with damage to the right hemisphere (RHD) with healthy controls on a humor processing task (Wapner et al., 1981; Brownell et al., 1983) In these studies, joke stems with different alternative endings were presented and patients and healthy controls were instructed to select the correct funny punchline. In an extension of these studies Bihrle et al. (1986) investigated the differential contribution of the left and right hemisphere to the two stages of incongruity resolution theory. In this study the verbal joke stimuli from Brownell et al. (1983), as well as captionless cartoons, were presented to RHD patients. In addition, the cartoons were presented to patients with damage to the left hemisphere (LHD). While both patient groups were impaired, they showed a differential pattern of errors. RHD patients mostly chose nonsensical endings that involved detection of an incongruent element, but no resolution of this incongruity. LHD patients were mainly attracted to straightforward endings that were congruent in the story context, but did not involve incongruity, that is, were not funny. The findings have therefore been interpreted as support for a differential contribution of the two hemispheres, with the right hemisphere being engaged in incongruity resolution and the left hemisphere supporting incongruity detection.

In a more recent investigation by Shammi and Stuss (1999), a paradigm similar to that of Brownell et al. (1983) was used. In contrast to previous studies, funniness ratings were also assessed. The results suggested a predominant role of the right frontal cortex in humor processing. Shammi and Stuss (1999) reported that patients with lesions to the right frontal cortex showed deficits in selecting the correct funny punchline. They showed reduced funniness ratings and typically selected nonsensical endings. In a recent functional imaging study (Goel & Dolan, 2001), activation in response to jokes rated as funny was compared to activation for nonfunny jokes on a participant-by-participant and joke-by-joke basis. Activations were observed in the medial ventral prefrontal cortex and bilaterally in the cerebellum. Since the activation of medial ventral prefrontal cortex also covaried with participants' post-scan ratings of joke funniness, Goel and Dolan (2001) concluded that this region is part of a network involved in processing the affective component of humor.

The “frontal aging hypothesis” suggests that the prefrontal cortex is particularly vulnerable to the effects of aging (e.g., West & Covell, 2001; Tisserand & Jolles, 2003) and implies that functions supported by the frontal lobes will be disproportionately impaired. This hypothesis is based on age-related alterations in the volume of the frontal cortex (Tisserand et al., 2002) and pronounced age-related decreases in 5-HT2 and D2 receptor availability in the frontal lobes (Wang et al., 1995). Impairments of executive functions such as working memory, inhibition, and planning have consistently been reported (e.g., Brennan et al., 1997; Milham et al., 2002; Thompson-Schill et al., 2002, Treitz et al., in press). Recent cognitive studies have suggested that executive tasks are typically impaired in normal aging, whereas social decision-making tasks are less commonly affected (MacPherson et al., 2002). The authors interpreted their findings in terms of greater impairment in functions supported by dorsolateral prefrontal cortex, with preservation of functions supported by ventromedial prefrontal cortex.

In an early study by Schaier and Cicirelli (1976) three age groups were compared. It was observed that comprehension of jokes decreased with age whereas appreciation increased. Shammi and Stuss (2003) compared older and younger groups, and reported that the older group showed impairment in a joke completion task that required the selecting of correct, funny punchlines from a set of alternatives. The commonest error of the older group included the straightforward and slapstick endings. In addition, performance correlated significantly with measures of executive functions. The authors concluded that comprehension of the cognitive component of humor, which is probably mediated by the dorsolateral cortex, may deteriorate with aging. On a separate task, participants were instructed to rate the funniness of short neutral and humorous verbal statements. The age groups did not differ significantly with respect to the difference score between humorous and neutral items. The authors interpreted the latter findings as evidence of intact (affective) humor appreciation with aging, and suggested that this might be linked to medial frontal structures. The conclusions of Shammi and Stuss (2003) are, however, limited by differences in the complexity of the different tasks used to assess the affective and cognitive components. In addition, it is possible that participants detected that an ending was the correct funny punchline, but still found another ending funnier (e.g., slapstick). Thus the assessment of funniness ratings for each of the different alternatives would be of relevance. Further studies are therefore desirable to shed light on the putative dissociation of the affective and cognitive components of humor in normal aging.

While the role of impaired executive functions in mediating humor processing deficits in normal aging has received some attention, the possible influence of mentalizing (theory of mind) ability has not been investigated. Mentalizing refers to the ability to reason about mental states, including beliefs and intentions, and may be mediated by medial frontal structures (Stuss et al., 2001). One study (Winner et al., 1998) examined the ability of RHD patients to distinguish lies from jokes, and concluded that performance on this task was correlated with the ability to make mental state attributions on a Theory of Mind (ToM) measure. The importance of mentalizing for humor processing was supported by a later study (Happé et al., 1999). Studies of humor processing in autism have also been interpreted in terms of a mentalizing contribution (Happé, 1994). Mentalizing may thus provide a separate route to humor processing, independent of executive skills. Studies investigating mentalizing in normal aging have produced mixed findings, with some suggesting it to be unaffected (e.g., Happé et al., 1998; Saltzman et al., 2000), and others indicating age-related decline (e.g., Maylor et al., 2002; Sullivan & Ruffman, 2004).

The aim of the present study was to investigate the influence of normal aging on cognitive and affective components of humor processing, and to examine the relationships with mentalizing and executive skills. For the assessment of humor processing a similar type of task as in the studies by Brownell et al. (1983) and Shammi and Stuss (1999, 2003) was used with the different alternatives reflecting incongruity and/or resolution. In an extension of previous studies, funniness and logical ratings were obtained for the different alternative punchlines.

METHOD

Research Participants

Eighty-seven healthy participants gave informed consent to take part. The study was completed in accordance with the Helsinki declaration. Participants were given a screening interview and excluded if they had a history of psychiatric or neurological illness, head trauma, or substance abuse. The Mini-Mental State Examination (Folstein et al., 1975) was also administered to exclude participants with possible dementia (cut-off score: 26).

Participants were assigned to three age groups: younger (20–34 yrs; n = 32; 12 males, 10 females), middle (40–58 yrs; n = 29; 10 males, 19 females), and older (60–78; n = 26; 12 males, 14 females). The number of males and females did not differ significantly (χ² = 0.84; p = .67). The three age groups did not differ significantly in intellectual ability (IQ), assessed by the subtests “Similarities” and “Picture Completion” of the reduced Wechsler Intelligence Scales (WIP; Dahl, 1986) or in mood, assessed by the Beck Depression Inventory (Beck & Steer, 1987). Demographic variables are shown in Table 1.

Demographic data, general intellectual functioning, executive functions, and affect (means and standard errors) in the three age groups

Humor processing and mentalizing

For the assessment of humor processing, a new computerized task was developed, similar to that used in the studies by Brownell et al. (1983) and Shammi and Stuss (1999, 2003), with the different alternatives reflecting incongruity and/or resolution. One hundred and thirty jokes were selected from a large pool, excluding those with potentially offensive themes including religion, politics, illness, or mortality. Each of these jokes was classified by two raters as being dependent on mentalizing ability for the comprehension of the joke. In a second stage, the two original raters and three additional raters assessed the funniness of each joke on a rating scale from 0 (“not funny”) to 3 (“very funny”). The raters were aged 47, 30, 25, 26, and 30 years with a mean age of 31.6 (SD = 8.90). Twenty-four of the jokes with a funniness rating of > 1 were included in the final test battery.

For each item, the joke stem was initially presented on the screen. In order to minimize memory demands, the joke stem remained on display throughout. After reading the joke stem, participants saw an array of four possible endings, and were instructed to select the correct funny punchline. Four types of endings were used: the correct funny punchline (C), a slapstick ending (S), an illogical ending (I), and a logical ending (L). After the selection of an ending, the joke stem was presented again with each of the alternative endings one after the other. Participants were then asked to rate the funniness (“How funny is this ending? Please rate how funny it makes the story”) and logic (“How logical is this ending? Please rate how well it fits to the story?”) for each alternative ending on a four-point scale (not funny, slightly funny, fairly funny, very funny and very badly, fairly badly, fairly well, very well). General comprehension was assessed by two simple factual (nonmentalistic) questions. Mentalizing ability was assessed by three mentalistic questions for each joke. During the presentation of the mentalizing questions, the joke including the correct funny punchline was shown on the screen. The first question referred to the perspective of protagonist 1, and the second to the perspective of protagonist 2. The third question referred to the comprehension of the correct, funny punchline. The participants' responses were recorded. The responses to the mentalizing questions were then scored as follows: 2 points for correct answers, 1 point for partially correct answers, 0 points for incorrect answers. All responses were rated by two independent raters, and inter-rater agreement was > 95% for all questions. Disagreements were judged by a third independent rater. The general comprehension and mentalizing questions were always presented in the same order.

Example of a Joke

Joke stem

“Martin had just started his own company. When a visitor came into the office, Martin picked up the telephone. He pretended to be discussing a multi-million-pound deal. Eventually he put the phone down and said to the visitor: ‘Can I help you?’”

Alternative endings

C: “The visitor said: ‘Yeah, I've come to connect up your telephone.’ ”

I: “The visitor said: ‘The colour of this wallpaper matches my tie.’ ”

S: “Martin's chair suddenly collapsed and he fell on the floor.”

L: “The visitor said: ‘Yes, I'm looking for a job in your new company.’ ”

Nonmentalistic fact questions

1. “Was the company started by Brad?”

2. “Did Martin work in an office?”

Mentalistic questions

1. “Why did Martin pick up the telephone and speak?”

2. “What did the visitor think when he heard Martin speaking into the telephone?”

3. “What did Martin think when the visitor said ‘Yeah, I've come to connect up your telephone?’”

Measures of Executive Function

Inhibition

Inhibition of a habitual response was measured by the Stroop test (Bäumler, 1985). The participants were told to read aloud color words (printed in black ink) as fast as possible (reading), to name the color of colored lines (naming), and to name the ink color of color words printed in an incongruent color, such as “RED” in green ink (interference). The time taken for each task was recorded.

Set shifting

Psychomotor speed and set shifting were measured by the Trail Making Test (Reitan, 1992). In the first subtest participants drew lines to connect numbers in ascending order. In the second subtest they were asked to alternate between numbers and letters, in ascending order (e.g., 1-A-2-B). Time taken was recorded.

Working memory

Working memory was assessed using a number-letter sequencing test similar to the Letter-Number Sequencing subtest of the Wechsler Memory Scale (Wechsler, 1997). Sequences of letters and numbers were read out by the experimenter (e.g., 2-L-4). Participants were instructed to first repeat the numbers in ascending order followed by the letters in alphabetical order (2-4-L). The number of correctly reproduced sequences was recorded.

RESULTS

Humor processing

Selection of endings from alternatives

Analyses of variance (ANOVAs) with post hoc Tukey tests showed significant group differences for the number of correct [F(2,84) = 14.79, p < .0001], slapstick [F(2,84) = 3.05; p = .05], and logical choices [F(2,84) = 17.25; p < .0001]. The older group selected fewer correct punchlines when compared to the other two groups (both p < .0001). This is illustrated in Figure 1. They selected a higher number of logical endings in comparison with the two other age groups (both p < .0001) and showed a tendency to select more slapstick endings when compared to the younger group (p = .07). Differences were confirmed with nonparametric procedures in light of the floor effects observed in some conditions. The data are summarized in Table 2. The results for the funniness ratings are also shown in Table 2.

Results [means and standard error of the means (SEMs) of % correct] of the three age groups in the humor processing task.

Percentage of alternative choices for number of correct control questions, funniness, and logical ratings (means and standard errors) in the three age groups

Statistical analyses of funniness ratings were restricted to trials on which correct alternatives were selected. ANOVAs yielded significant age effects for the funniness ratings for the correct alternatives [F(2,84) = 4.75, p = .01], which was due to lower scores for the older group in comparison with the middle group (p = .008). All groups rated the correct funny punchline as funnier than the other three alternatives (all p < .0001). To obtain a clearer measure of the ability to differentiate between funny and neutral items, difference scores between the funniness ratings of the correct and logical alternatives were also calculated. ANOVA yielded significant differences [F(2,84) = 5.94, p = 004], which was due to lower scores for the older group in comparison with both other groups (both p < .02).

The results for the logical ratings are shown in Table 2. Statistical analyses of logical ratings were restricted to trials on which correct alternatives were selected. Analyses yielded significant effects for correct [F(2,84) = 3.30, p = .04] and logical endings [F(2,84) = 7.56, p = .001]. Older people rated logical endings as being less logical than both other groups (both p < .005). In addition, they showed a tendency to rate correct endings as being less logical than both other groups (both p < .07). To obtain a clearer measure of the ability to differentiate between logical and illogical items, difference scores between the logical ratings of the logical and illogical alternatives were also analyzed. ANOVA yielded significant differences [F(2,84) = 9.82, p < 0001], which were due to lower scores for the older group in comparison with both other groups (both p < .001).

The results for the nonmentalistic questions are shown in Table 2. ANOVA showed no significant differences between the three groups in the number of correct nonmentalistic questions [F(2,84) = 0.80, p = .45).

The responses for the mentalistic questions are illustrated in Figure 2. ANOVA yielded a significant group difference [F(2,84) = 14.22, p < .0001], which was due to significantly lower scores for the older group in comparison with both other age groups (both p < .0001). No other significant results were observed (p = .67). Separate analyses for the answers to the first two mentalistic questions (referring to the perspectives of the two protagonists) and the more complex third question (referring to the comprehension of the correct funny punchline) also yielded significant effects (both p < .002), which were due to significantly lower scores for the older age group in comparison with the younger groups.

Results [means and standard error of the means (SEMs)] of the three age groups in the mentalizing task.