UCL Deafness Cognition and Language Research Centre


DCAL responds to Harry Knoors weblog

6 November 2012

In June 2012, Harry Knoors (Professor in Deaf Education, Radboud University Nijmegen) blogged about research recently published by DCAL on age of acquisition effects in BSL.

The original blog was in Dutch; an English translation is below.

DCAL responded to Knoors' blog in English (for DCAL reply in English, click here and scroll down).

Harry Knoors then commented on DCAL's reply in Dutch; an English translation is below.

English translation of original weblog:

Harry Knoors: "Early sign language exposure"

Translation by Pamela Perniss and approved by Harry Knoors

"Early exposure to sign language helps Deaf children, research says." That is the headline on a British Deaf Society webpage. I came across it last week. With my curiosity piqued, I read on. Research is said to have shown that deaf children find learning new languages easy. Provided that they were exposed to sign language before the age of two. Because, as the article  continues, deaf children who are exposed to sign language later, like between the ages of 2 and 8, do notably worse. The website then quotes one of the researchers, Dr. Adam Schembri, giving support of this. "The advantages of early sign language exposure remain clear even with rapid advances in hearing aids and cochlear implants", asserts Schembri.

It just so happens that I am familiar with the research on which this report is based. Excellent research. Recently published in the leading scientific journal Cognition. Tells a slightly different story really. In particular, a much more complicated story. Schembri's group investigated 30 British deaf adults. Looked at their skills in judging the grammaticality of utterances in British Sign Language. And looked at their reading skills. Nine of the deaf adults were exposed to sign language from birth, they are native signers. They all had at least one deaf parent. Eleven of the deaf adults were exposed to sign language between the ages of 2 and 8 years, on average when they were 4 years old. The remaining ten were late signers; they came into contact with BSL between the ages of 8 and 18 years, prior to this communication with them had been only in spoken English or Sign Supported English.

The group of deaf who were exposed to sign language between the ages of 2 and 8 performed worst in all respects. Their BSL skills were rather disappointing and they read less well than both other groups. Remarkably, the native signers and the late signers perform equally well on the BSL task. And the late signers are the best readers, though less skilled than hearing adults.

Surprising results. You have to let them sink in a bit. Because what do they really mean? That exposure to sign language does indeed promote language and reading development. Provided the exposure is as early as possible, preferably before the age of 2. But also that late exposure to sign language doesn't necessarily have to be detrimental. Provided that the deaf individuals in question have been able to acquire a spoken language well. Because, in that case, they are then subsequently able to successfully learn a sign language as a second language. That is the most probable explanation, anyway, according to Schembri and colleagues, for why the deaf adults with late exposure to sign language performed best, to everyone's surprise…

Schembri and the other researchers hasten to declare that their research does not imply that bilingual education of deaf children should be given up. In addition, not all bets should be placed on learning sign language at a later age. Because how would you select the deaf individuals who would succeed at it? According to Schembri and colleagues it is necessary to make sure that deaf children get sign language exposure when they are babies and toddlers. How? Well, they don't go into details about that.

Because what are the chances that the primarily hearing parents [of deaf children] are able to quickly pick up a sign language? Sure, some are able to. But many others are not successful, in spite of obvious efforts. Not to worry, I hear some hearing and deaf colleagues say. It's just a question of bringing sign language users, deaf adults into the family as soon/early as possible. And indeed, that may be a sensible option. But certainly one [an option] that deeply intrudes on a family. Because it's not just a matter of having a deaf person at the table for an hour… In reality, you're sharing the responsibility of raising your child with a deaf adult. Quite a lot to accept/swallow, isn't it? No matter how nice and competent the deaf adult is. May be the best option, after all, for situations where the likelihood of successful communication in Sign Supported Dutch is small. For situations where the successful acquisition of spoken Dutch by means of a CI wouldn't go well. For situations where the successful acquisition of NGT [Sign Language of the Netherlands] as a second language would be pretty much impossible. In short, doesn't the research by Schembri and colleagues actually support calls for flexibility and differentiation in implementing policies of sign language exposure for deaf children.

English translation of Knoors' reply to DCAL:

Translation by Pamela Perniss and approved by Harry Knoors

It is good for the reader to be aware that my weblog referred to the interpretation of research by Kearsy Cormier and colleagues on a website from the British Deaf Society. It was within that framework that I went into some more detail about their research. The researchers are mistaken when they say that "Knoors states that the deaf adults with late exposure to SL performed best on the GJT." In my blog it says that the native signers and the late signers do equally well on the BSL task, exactly as it is stated in the article. Admittedly, this is the case for accuracy, but indeed not for reaction time. Reaction time is in fact slower for late signers, which by the way is typical of people using a second language. So, I don't attach too much weight to this. The core of the reaction by Cormier and colleagues lies in their observation that I maintain that late exposure to sign language does not need be detrimental, provided that the deaf in question have been able to learn a spoken language well. I agree with Cormier and her colleagues that the PROVIDED THAT bit should not be taken lightly. It is not given/self-evident that deaf people will be able to acquire spoken language as a first language, definitely not in the past, but even now that we are implanting deaf children early it's far from possible for all children. I don't claim that it is anywhere by the way. What I do say is that the risk that early exposure to spoken language for deaf children with a CI can bring must be weighed against the chance of being able to successfully teach these children sign language in their first two years of life.