Do the children with cochlear implants comprehend verbal irony?

Introduction: Despite great variability, many children with cochlear implants (CI) reach a linguistic competence comparable to normal hearing (NH) peers, even if prosody recognition remains a weakness. CI children have nonverbal mental age (NVMA) in line with normal hearing (NH) peers, but their Theory of Mind (ToM) abilities, such as irony comprehension are reported to be delayed. Their pragmatic abilities are understudied, and irony comprehension is a particularly interesting topic because it is not yet clear what are the factors that better predict the development of this ability in NH children: however, some scholars highlighted the role played by 2ndorder ToM abilities. Among ToM abilities, aim of this study was to assess irony comprehension in a group of children with CI and in two control groups of NH children. Primary goal was to evaluate their pragmatic abilities and identifying the factors that predict irony understanding. Methods: We tested a group of 28 children with CI (mean age 8 y and 5 months; range 3-12 years); 13 children had received the cochlear implant within 18 months of age (early CI), 15 after 24 months of age (late CI). We had also two control groups of NH children, one matched for chronological age (CA-NH, p =.905); and one for hearing age (HA-NH, p.=892). Children were administered Raven CPM for their NVMA; a morphosyntactic task for linguistic abilities; six tasks for ToM abilities (up to 2ndorder ToM); a task for prosody recognition; a new irony comprehension task that included 6 ironic and 4 literal remarks. Results: All the CI children had lower linguistic abilities compared to the CA-NH (p=.027), but not with HA-NH (p=.614). The same level of ToM was reached by the CI and HA-NH groups, whereas a statistically significant difference was observed between CI and CA-NH in 2ndorder ToM scores (p=.026). Prosody recognition was also significantly worse (p<.01) in all the CI recipients with regard to both NH controls. On the irony comprehension task, CI children had a lower accuracy in ironic remarks not only compared to CA-NH (CI: 52% vs CA-NH: 82; p<.001), but also with regards to HA-NH (CI: 52% vs CA-NH: 73%; p=.011), while literal comments were at ceiling for all groups. The comprehension of ironic criticisms strongly correlated with hearing age, and with ToM skills. Children with early CI had higher level of accuracy in irony comprehension than those in the late CI. Conclusion: Irony comprehension in children with CI is delayed compared to that observed in the NH peers, also in the younger ones matched for hearing age. This deficit does not seem to depend on their difficulty in prosody recognition, but on their delayed ToM skills.