Artificial hearing, natural speech: Cochlear implants, speech production, and the expectations of a high-tech society (review)

Cochlear implants (CIs) represent the first successful attempt to replace a human sense with electronic means. These devices are primarily indicated for patients whose inner hair cells (the neurons that convert sound into neural information) do not function. In these cases, amplifying sound with a conventional hearing aid does not work because the crucial step of converting sound to neural activity fails, no matter what the intensity of the sound. CIs bypass the normal hearing mechanism by stimulating auditory neurons electrically, via electrodes implanted in the inner ear. An external part includes a microphone, a speech processor that encodes sound information into electrical stimulation patterns, and a transmitter antenna. This antenna sends the stimulation information to an implanted receiver that decodes the information and sends stimulation pulses to electrodes placed in the inner ear, in the snail-shaped organ named cochlea. The first CIs were built and tested in humans by a French team in the late 1950s. Since then, after a process that included many breakthroughs as well as some setbacks, CIs have become the standard of care for treating many kinds of deafness. As of this writing, more than 67,000 patients in the United States and close to 200,000 worldwide have received CIs. Most of these patients can understand at least some speech without the aid of lipreading, and even congenitally deaf children can develop speech perception, speech production, and oral language skills that are sufficient to allow them to attend mainstream schools.

Despite the fact that results are very good on average, a major hallmark of CI outcomes is their individual variability. Each group of patients, no matter how uniform their demographics might be, tends to include ‘stars’ as well as more mediocre performers. There are important differences between groups as well, and it is useful to think about three different groups when discussing the scientific interest and the clinical impact of CIs. The first group includes congenitally or prelingually deaf children. When they are implanted in the first few years of life, the CI not only helps these children’s speech perception, but it also facilitates their development of intelligible speech production and that of language skills in an oral language such as English. Most of these children obtain significant benefit from their CIs, can understand at least some speech without lipreading, and can participate in mainstream classrooms. A second group is composed of congenitally or prelingually deaf individuals who are implanted in their adolescence or adulthood. Communicative benefit is much more limited than in the first group. Patients are able to detect sounds and obtain some lipreading enhancement, but speech perception is generally limited. Speech production and oral language skills are also much more limited than in the first group. The third group—and the one that is the focus of Lowenstein’s book—is that of postlingually deafened adults, people who developed oral language and speech production prior to becoming deaf. Despite years or even decades of profound deafness, changes in speech production are subtle. By and large, patients in this group have no major difficulty producing intelligible speech. The main benefit these patients obtain is improved speech perception. Using their CIs, the vast majority of them are able to have fluent face-to-face conversations, and most of them can communicate on the telephone.

Beyond the clinical benefit they provide, patients with CIs are also interesting because they may help answer basic questions about human communication, such as those concerning the role of hearing on speech production. Postlingually deaf CI users undergo several years of sound deprivation and then start receiving auditory input again (albeit a degraded version) once they receive CIs. The subtle changes in speech production before and after implantation in this population have been studied before, and much of the literature is reviewed in L’s book, focusing on a...