Wie Sprache entsteht: Spracherwerb bei Kindern mit beeintrachtigtem und normalem Horen (review)

Given the politically charged atmosphere surrounding the use of cochlear implants in deaf children, Gisela Szagun’s longitudinal study (‘The emergence of language: Language acquisition by children with impaired and normal hearing’) of spoken German development in twenty-two deaf children with cochlear implants (CI) compared to twenty-two hearing children is refreshingly straightforward. Josh Aronson’s award-winning documentary Sound and fury (2000) introduced many people to the controversial issues of identity and language surrounding CIs. S’s study moves this controversy forward by providing a meticulous analysis of theoretically relevant measures of language development in a sample of deaf children with CIs who are representative of the broad range of possible outcomes. The results of S’s study will not be welcomed with open arms by staunch supporters of either side of the CI debate. CIs are not the ‘wonder cure’ of deafness (for those inclined to seek one), but in limited circumstances, children with CIs can obtain impressive levels of spoken language competence. This book is not written solely for those with a preformed opinion about cochlear implants. Anyone interested in the degree to which language acquisition is data-driven will want to know how language develops in this population of children who have only partial access to auditory input. S’s clear writing and her willingness to take a firm position on controversial issues make the book fascinating reading for linguists who have never encountered a deaf child, as well as for the medical and educational professionals who work with deaf children every day.

To summarize the results very coarsely, language development in this population of twenty-two CI children, who were implanted between the ages of 1;2 (years;months) and 3;10, was significantly slower than in a population of twenty-two hearing children, even when matched for vocabulary level—not age—at the outset of the study. In other words, at the start of the study, all of the children had a mean length of utterance (MLU) of 1 (CI: 1.04, hearing: 1.05) and an expressive vocabulary of about twenty words (CI: 21, hearing: 18). The CI children as a group lag behind the hearing children on measures of vocabulary development, MLU, verb inflection, and the use of definite articles (which encode gender and case in German). But this delay is apparent only when the CI children are treated as a single population, which they clearly are not. Neither, for that matter, are the hearing children. The twenty-two hearing children fell into two fairly even groups of thirteen ‘fast’ learners (H1) and nine ‘slow’ learners (H2). Three of the deaf children (CI1) were on par with the H1 group. Seven of the deaf children (CI2) developed along a similar time course as the hearing ‘slow’ learners. The remaining twelve deaf children did not perform nearly as well as the hearing group, with six of those children making almost no gains at all during the first eighteen months of the study. Unfortunately, the differences between the faster and slower language learners only increase over the course of the study.

S carries out a series of correlations to determine the influence of preoperative hearing level, postoperative hearing level, and age of implantation on language development. Both pre- and postoperative hearing levels are highly correlated with MLU, but surprisingly, age of implantation is not. The effects of age of implantation may be difficult to detect because the CI did not influence language development for all children. Specifically, deaf children with the mildest hearing losses (50–65 dB) developed language like one of the hearing groups irrespective of age of implantation. These children were receiving some auditory input to spoken German prior to implantation. It’s possible that these children would have developed language on a typical timetable without the implant; alternatively, preimplantation auditory stimulation may have developed the neural systems necessary for processing speech input...