Difficulties in understanding irony and sarcasm are part of the social cognition deficits in patients with schizophrenia. A number of studies have reported higher error rates during comprehension in patients with schizophrenia. However, the relationships of these impairments to schizotypal personality traits and other language deficits, such as the comprehension of proverbs, are unclear. We investigated irony and proverb comprehension in an all-female sample of 20 schizophrenia patients and 27 matched controls. Subjects indicated if a statement was intended to be ironic, literal, or meaningless and furthermore rated the meanness and funniness of the stimuli and certainty of their decision. Patients made significantly more errors than controls did. Globally, there were no overall differences in the ratings. However, patients rated the subgroup of stimuli with answers given incorrectly as having significantly less meanness and in case of an error indicated a significantly higher certainty than controls. Across all of the study participants, performances in irony (
Defective appraisal of the intention of others and difficulties with language are hallmark features of psychopathology in schizophrenia. Social cognition deficits have been previously identified and are currently being extensively researched. The results show that deficits in social cognition are relevant to real-world functioning and outcome [
This is obvious, considering that the decision on whether a remark made by others is intended to be ironic or not is slightly artificial, but is instead required routinely in everyday interaction. Everyone is familiar with irony. It is remarkably frequently used, as shown by linguistic research [
Previous research has shown that patients with schizophrenia, as a group, have difficulties in irony comprehension [
However, defective irony comprehension in schizophrenia is likewise influenced by numerous additional cognitive operations that individuals with schizophrenia are known to have difficulties with. For example, irony also relates to higher-level language comprehension [
Another interesting point that has not received much attention in past research is how patients with schizophrenia appreciate ironic remarks. Most previous studies have addressed performance (i.e., whether patients fail to comprehend ironic remarks), while it is widely unknown if patients with schizophrenia “like” ironic remarks or they perhaps judge their intention as more mean than healthy controls would do. This relationship to the perceived “meanness” by patients with schizophrenia seems particularly interesting, because judging the intention of others as “mean” or “hostile” may relate to the origin of delusions in these patients [
There is consensus in the literature that the severity of irony comprehension deficit differs between individuals with schizophrenia. However, it is currently unclear whether this deficit depends on a continuum with schizophrenia and what the possible mediating factors could be. One possible mediating factor for deficits in sarcasm comprehension that has been discussed, but not consistently found, in the literature is schizotypal personality traits [
Our pilot study aimed to investigate both comprehension and appreciation of ironic remarks in a sample of patients with schizophrenia. On the one hand, we aimed to investigate comprehension and whether it was related to comprehension of proverbs, the most often applied measurement of concretism in schizophrenia. We applied an irony comprehension task recently developed by our group [
We have several hypotheses concerning the results. Since lower performance in irony detection is documented in several previous investigations [
20 right-handed [
Clinical and sociodemographic characteristics of patient and control groups.
Healthy controls | Schizophrenia | Significance | |
---|---|---|---|
|
|
||
Age | 38,9 | 34,1 | n.s. |
Education years1 | 15,1 | 14,2 | n.s. |
Digit span (digits) | 6,2 | 6,1 | n.s. |
CPT errors | 0,1 | 1,5 | 0.0026 |
Verbal intelligence3 | 31,3 | 30,6 | n.s. |
HAWIE picture sequencing test2 | 32,5 | 26,4 | n.s. |
SPQ cognitive perceptual4 | 10,1 | 22,3 | 0.00007 |
SPQ interpersonal4 | 5,7 | 14,8 | 0.00005 |
SPQ total score | 15,6 | 36,7 | 0.00000 |
PANSS total score | 70,3 | ||
SANS total score | 36,1 | ||
SAPS total score | 35,7 | ||
Global assessment of functioning scale | 37,9 |
1In full-time education.
2Subtest 2 from [
3Multiple-choice vocabulary test [
4As defined by [
Patients were recruited from the Department of Psychiatry, University of Tübingen, Germany. Patients were acute or subacute inpatients and predominantly of the paranoid subtype. All participants were native German speakers, had no other past or present medical illness, and had sufficient reading skills. A subgroup of the sample additionally participated in a functional magnetic resonance imaging study [
First, all subjects received complete information about the study and ability to consent was ensured. Then, subjects underwent a practice session for the irony comprehension task with stimuli not used in the experiment and provided written informed consent. The study was approved by the local ethical committee (University of Tuebingen, Germany). Subjects then completed the irony comprehension task. The sequence of the other tests was counterbalanced between the subjects. If requested by the subject, a short break between the tests was possible.
An irony comprehension and evaluation task was applied to all subjects. Stimuli and stimulus sequence are identical with a previous functional magnetic resonance imaging study [
The irony comprehension task consists of 56 short German text vignettes as stimuli, each consisting of an introductive context (2 sentences, 8–12 words) and two or more protagonists (e.g., “Petra hates fish. Her mother has cooked salmon for her.”). The introductive context is then followed by a target part, in which one of the protagonists makes a statement (e.g., “Petra says: ‘Oh brilliant, my favourite meal!’”). This statement has, in this context, either an ironic or a literal meaning or is meaningless.
The number of words and sentences, grammatical complexity, and word frequency are counterbalanced between literal, ironic, and meaningless context scenarios.
Corresponding literal and ironic target sentences are identical. For example, for the above ironic stimulus, the literal counterpart would be, “Katja loves spaghetti. Her mother has cooked a lot of spaghetti for her. Katja says: ‘Oh brilliant, my favourite meal!”’
A comprehensive description of the stimuli and their development and evaluation process is given in [
During the irony task, subjects were seated in front of a notebook computer (Dell Inspiron) next to one of the investigators (K.L.). In-house software was used to present the stimuli and measure the button press response. Stimuli were presented in a pseudorandomized order not foreseeable for the subject. The introductive context was both presented visually and read out by the investigator. To avoid any effect of prosody, target sentences were only presented visually. The task was to indicate by pressing one out of three buttons whether the target sentence was in this context most likely ironic, literally meaningful, or meaningless. After each stimulus and button press, subjects were asked by the investigator (K.L.) to rate the certainty of their decision on a 5-point Likert scale (from 0, very uncertain, to 4, very certain) as well as the meanness (from 0, not mean at all, to 4, very hurtful) and jocularity (from 0, not funny at all, to 4, very funny) of the protagonist’s statement.
A multiple-choice German proverb test developed at the University of Münster [
Several scores were defined in our investigation. First a proverb familiarity score was calculated for each individual. It consists of the mean values of the familiarity ratings of all 32 proverbs. Secondly, mean values for numbers of proverbs rated as unfamiliar were compared between patients and controls. Here, we defined a proverb as “familiar” if its rating is above 1; that means it is ready counted as familiar if the subject indicates he has heard it once before. Each Type I answer given is counted as one point, whereas all other answers in multiple choice are counted zero. This results in a sum score for each participant. The higher the score is, the better the performance is. The highest reachable sum score is 32. All these scores were compared between the groups.
After completion of the previous tests and the schizotypal personality questionnaire [
In addition, patients were interviewed by a psychiatrist (A.R.) to evaluate current psychopathology using the SAPS [
SPSS 11.5 was used for all statistical analyses. Due to the exploratory character of the study and the small sample size, a significance level of 0.05 was chosen.
Pattern of errors in the irony comprehension test of patients (left) and control subjects (right).
For the whole task with all of the stimuli taken into account, the patient and control groups did not differ in their certainty, in their judgement of meanness, or in how funny they rated the stimuli (Table
Group differences in detection and appreciation of irony.
Healthy controls | Schizophrenia | Significance | ||
---|---|---|---|---|
All stimuli | Certainty | 3,81 | 3,73 | 0.26 |
Meanness | 0,58 | 0,52 | 0.58 | |
Funniness | 0,57 | 0,51 | 0.54 | |
|
||||
Literal stimuli | Certainty | 3,81 | 3,85 | 0.59 |
Meanness | 0,09 | 0,15 | 0.26 | |
Funniness | 0,22 | 0,15 | 0.43 | |
|
||||
Ironic stimuli | Certainty | 3,79 | 3,67 | 0.21 |
Meanness | 1,02 | 0,77 | 0.24 | |
Funniness | 1,09 | 0,76 | 0.10 | |
|
||||
Meaningless stimuli | Certainty | 3,88 | 3,60 |
|
Meanness | 0,70 | 0,69 | 0.95 | |
Funniness | 0,19 | 0,70 | 0.002 | |
|
||||
Correct answers | Certainty | 3,84 | 3,82 | 0.70 |
Meanness | 0,56 | 0,49 | 0.55 | |
Funniness | 0,57 | 0,41 | 0.16 | |
|
||||
Incorrect answers | Certainty | 2,34 | 3,34 |
|
Meanness | 2,23 | 0,72 |
|
|
Funniness | 0,52 | 0,60 | 0.64 |
In a next step, the ratings for answers given correctly and the answers given incorrectly were compared between the groups. Control subjects and patients with schizophrenia reported significantly lower certainty for the stimuli that they responded to with an error compared to those to which they replied correctly (
Detection and appreciation of irony. Group differences.
In the case of meanness ratings, control subjects (
There were no differences between correct and incorrect stimuli in either group for perceived funniness.
This analysis was calculated for all participants together, following the rationale that schizotypal traits represent a continuum [
Correlation between performance in the (a) irony comprehension test and (b) proverb comprehension test and schizotypal personality traits.
We investigated the comprehension and appraisal of higher-level language in female patients with schizophrenia and matched healthy control subjects. Subjects completed a German irony comprehension and appraisal test [
Beyond investigating performance, a main goal of the study was to understand how patients with schizophrenia would evaluate and appreciate ironic remarks. As outlined in the introduction, we hypothesized that patients with schizophrenia would be less certain in their decision about intention and would indicate a higher degree of meanness for the stimuli, especially those intended to be ironic, and would rate them less funny. None of these hypotheses was clearly confirmed. Only meaningless stimuli were rated significantly higher in their funniness, a finding that is compatible with previous findings of differences in funniness ratings in patients with schizophrenia [
An additional aspect to our investigation was the evaluation of performance in the multiple-choice proverbs test. As expected, we found an impairment in schizophrenia. Our finding of an impairment of proverb comprehension in schizophrenia mirrors numerous classical findings [
Irrespective of the diagnosis when means were correlated across all study participants (patients and controls), the performance was correlated with schizotypal personality traits: the higher the total score of the schizotypal personality questionnaire (SPQ), the lower the performance in either the irony or proverb test. This result is in line with increasing evidence suggesting that schizotypal personality traits represent a continuum from healthiness to schizophrenia [
Another limitation is that patients and control subjects significantly differ in their expression of schizotypal personality traits, although Figure
We are aware of several limitations in our study. Several of them are related to the characteristics of our irony comprehension task [
Sample size is another limitation of this study. Although it was within the range of comparable investigations [
In this study, we found a significant association of schizotypal personality traits with performance. However, SPQ values were significantly higher in patients with schizophrenia, and in our pilot study, we did not control for other personality variables. For example, other personality traits not measured here may be associated with irony comprehension as well [
Another limiting factor in our study is gender. To enhance homogeneity of the sample, we investigated only female subjects. This point is relevant since gender differences have been reported for irony appreciation in healthy subjects [
We see several implications from our study. Similar to a previous study [
Despite these limitations, we were able to demonstrate that female patients with schizophrenia are impaired in irony comprehension and show reduced insight and aberrant response, especially to those stimuli judged wrong. This, together with the observation that irony is so remarkably frequent in everyday conversations and is even more frequent in difficult communicative situations, which per se have a higher ambiguity, makes irony an especially promising candidate for social cognition training in patients.
The authors declare that there is no conflict of interests regarding the publication of this paper.