The United Nations says all people have the right to a language. But some people in the Cochlear Implant industry are prepared to gamble with that right. The gamble is that a Cochlear Implant (CI) will be successful for the child. It’s a gamble because if the part of the brain that deals with language is not developed, then the child may never acquire a language, whether it be sign language or spoken language. There is only one way to develop the language part of the brain – via exposure to a proper language at a young age. Deaf people say “Many audiologists have a medical view. They try to fix the ears and neglect sign language.“
A Cochlear Implant (CI) is a surgically implanted electronic device PLUS intensive speech therapy, which is why the videos of a child being “switched on” are misleading. The implant is connected to the auditory nerve then the skin is sewn up and usually heals fine. An external wire goes from a hearing aid behind the ear to the head and attaches magnetically through the skull to the implant – so CAT scans are very dangerous. A CI kid needs continuous audiology appointments and ongoing intensive Auditory-Verbal Therapy. A CI needs enormous dedication and commitment from the family. It’s not a magic bullet. The CI industry also acknowledges that CI does not work in many cases.
The CI industry often says that sign language for CI kids is not needed and may even stop them learning a spoken language. Research shows this is not true. Figures are hard to get, but an estimated 10% of CI kids cannot speak as well as their peers by age 5 and many have no language at all. The results are better if a kid has a CI for both ears which are implanted at the same time. The CI industry often says if the CI doesn’t work THEN the kids can learn sign language. This is an irresponsible gamble.
So I was very interested to go to audiologist Julia Sarant’s talk (pdf) on Tuesday 9 August 2016 at the Hearing House in Auckland, New Zealand about these issues. The Hearing House paid for her meta study and the task was “to compare Oral Communication with Oral Communication plus Sign”. In her meta study, Oral Communication (OC) means teaching a CI kid a spoken language. As we will see below, it’s not clear from her meta study if “Sign” means a natural sign language or some other form of manual communication such as Signed English. I met Julia over a cup of tea afterwards. She said she couldn’t tell me who wrote the poor studies which she included in her meta study because it was important to maintain good relations in the industry. She really feels for the grief that some hearing parents never recover from when a Deaf child is born to them. Often that grief harms the child as in this sad subtitled documentary where a 4 years old doesn’t know their name.
Many worried mums were at the talk and a few dads and some hearing and education professionals. The questions after the talk showed many parents were confused and worried about the choices they had made or would soon have to make for their kids. They looked to Julia and her meta study for reassurance. It was also clear that one or two of the professionals did not know how to evaluate a meta study.
A meta study is not new research. It evaluates previous studies to try to find the truth (despite what postmodernists believe, truth exists). The conclusion of a meta study depends on what studies are included. A good meta study will try to cover the field. My preference is to avoid including poor studies in a meta study. However, it might report them in an introductory literature review or use them as examples of a failure.
Julia concluded, perhaps wrongly as we will see, that for Deaf kids “There is no evidence that bilingual education [OC plus sign language] improves educational outcomes over oral education.” Her meta study (pdf) looked at 39 other studies, but only 32 are listed in her bibliography (and one of these is an outdated 2001 paper by Harris et. al. about how to rank studies when she should have used the current USPSTF Quality of Evidence guidelines). So we do not even know what the 39 studies were. She said they covered six areas:
- speech perception
- speech production
- language development
- social development.
Julia rated one study as “good” for one of these six areas. All the rest were fair or poor for all categories. That result calls into question why other studies were not chosen by Julia to be part of her meta study. It is worth quoting in full from the Gallaudet University Linguistics Department which lists some of those studies:
To properly evaluate the effect of ASL [American Sign Language] on early English development, the appropriate comparison to make is between children raised with oral communication only and children raised with spoken English plus a full, natural sign language. Studies making this comparison have been appearing at an increasing rate, focusing on deaf cochlear implant users raised by deaf, signing parents. This population is relatively small but internationally studied. Recent research from Belgium (Mouvet 2013), Brazil (Quadros, Cruz, & Pizzio 2012), Iran (Hassanzadeh 2012), Italy (Rinaldi & Caselli 2009, 2014), the Netherlands (Giezen 2011, Giezen, Baker, & Escudero 2014), and the United States (Davidson et al. 2014) all report successful spoken language development for deaf cochlear implant users who are raised with natural sign language input. All of these studies have come to the same conclusion: There is no evidence that early exposure to a full, natural sign language obstructs spoken language development. In fact, Marcel Giezen (2011) concludes in his dissertation by saying that the evidence points to bimodal bilingual training as the optimal choice for cochlear implanted children, giving them the best chances for language success. (Source: Gallaudet University Linguistics Department)
Why did Julia not mention a single one of these studies? When I asked her later via email which was the “good” study which she included in her meta study she said it was Teresa Ching: Ching, T. Y. C. (2015). Is Early Intervention Effective in Improving Spoken Language Outcomes of Children With Congenital Hearing Loss? American Journal of Audiology, 24(3), 345–348. http://doi.org/10.1044/2015_AJA-15-0007. Again, Teresa’s study is not in the bibliography of Julia’s paper. This is another oversight. I eventually got a copy of Teresa’s short paper, but have been unable to get a copy of the data on which it relies. It’s unclear from Teresa’s study how many CI kids also learnt sign language. When I emailed Teresa’s office and left two messages to clarify this, I got no response, though they had earlier emailed me the information which helped me track down Teresa’s study. Elizabeth Levesque presented a slide (below) about Teresa’s study which is known as the Longitudinal Outcomes of Children with Hearing Impairment (LOCHI). Elizabeth discussed the LOCHI data during her 2016 presentation to the Australian and New Zealand Conference for Educators of the Deaf (ANZCED). The slide also shows that the reported data from Terera’s study is unclear. Why? Because we don’t know how many of the 22% of LOCHI kids were taught natural sign language rather than some form of Signed English.
So I am not convinced by the reliance Julia makes on Teresa’s paper to conclude that oral communication is better than oral communication plus sign language for language development.
Besides the studies above which Gallaudet reports, the Oxford Handbook of Deaf Studies, Language and Education also contains a chapter by Rachel Mayberry reporting the opposite of what Julia claims. Mayberry showed that Deaf children who learn to sign before they learn English can later learn to read and write English with native proficiency. But if Deaf children are taught spoken English as a first language but fail to learn it properly, then their chance of ever learning any language properly, whether sign language, oral language, or written language is zero. Mayberry also discussed two related and critically important points. First, language structures in the anterior left hemisphere of the brains of Deaf and hearing people are the same. Second, if these structures are not activated properly (in the early critical period for acquiring languages) then the child can never learn any language properly.
So in Julia’s meta study she asks us to believe in one study, and one study only, which she rated good and says supports her case. She does not give enough information for us to judge the worth of this study and she does not mention the other studies I have referred to here. This is not the way science proceeds. Science progresses by the whole scientific community reaching a consensus on all of the best available data as Naomi Oreskes shows in this entertaining and informative video. We need to share data, we need to know the details and we need to be able to test the claims for ourselves. Unfortunately I would rate Julia’s paper poor. It does not give any evidence, other than her rating of Teresa’s study.
A parent said to Julia during question time after the talk, “Your message is a great relief to me”. I am almost in tears as I write this because Julia said “There is no evidence at all that bilingualism helps.” I wrote down these words in my notebook as she said them. The words actually contradict her paper which said “There is limited evidence that early use of sign language may assist with the development of spoken language.” I’m not sure whether by ‘limited’ Julia means ‘some’ (in a positive sense) or ‘very little’ (in a negative sense’). In either sense, Gallaudet and Oxford Handbook contradicts this. There is lots of evidence that bilingualism (having both sign language and spoken language) can benefit both languages. When asked by someone else in the audience about the evidence from Colorado which also contradicts it, Julia she said “I think it was one study.”
If parents were properly informed I don’t think they would take the risk of not teaching their CI kids sign language. It is against the human rights of their child to gamble with language acquisition. Unfortunately by the time you discover your CI kid is behind in spoken language, it is likely to be too late. There are also heaps of stories on the internet from CI people who speak well but also wish they were taught sign language. My call to action is to encourage parents to form strong bonds with the Deaf community and to have a fluent signer interact regularly with their child. One of the big myths is that sign language is hard to learn. I’m hoping to blog about that myth soon.
This all fills me with sadness for the kids who are missing out.
(I’d like to thank the friends and members of the Deaf community who made very useful comments on earlier drafts of this blog.)