I ran across this thread on AllDeaf: http://www.alldeaf.com/sign-language-oralism/62109-some-thoughts.html

After quoting a study:

In summary, higher speech intelligibility scores acquired in 8- to 9-year-old congenitally deafened cochlear implant recipients were associated with educational settings that emphasize oral communication development and placement with hearing peers. Less accurate speech intelligibility scores at 8 to 9 years of age were associated with educational programs that emphasize the development of language via signs and placement in special education classes. (Tobey,  Rekart, Buckley,  Geers, 2004)

The user Faire Jour starts the thread with, “I think this would be in direct opposition to the idea that signing has no effect on speech development.”

This is largely a straw man argument.  There are two problems here. The first problem is that the study is only about one aspect of speech development: intelligibility. However, the study’s introduction refers to other studies that show that intelligibility is a major factor in effective oral communication. The second problem, the more important issue, is that most advocates for ASL are not saying that signing has no effect on speech development.

In all of these discussions we really need to be clear about what we want to understand. If all we care about is whether children with cochlear implants can pronounce words intelligibly, this is the study for us. But the debate is rarely centered on this one specific topic.

Before the advent of cochlear implants, this same argument against signing was made– that signing with deaf children would prevent them from developing speech. There are some competing theories here. One theory states generally that signing with deaf children has a net negative impact and that the best approach is to use only spoken language with deaf children (“oralism”). The other theory states generally that signing with deaf children has a net positive impact and that in general deaf children should be exposed to as much ASL as possible as early as possible (“bicultural/bilingual”). These are the two main groups, really, and there are variations within each camp. But, ultimately it boils down to whether signing is good or bad.

So with that in mind, let’s look further at what the study says. From the summary:

In summary, higher speech intelligibility scores acquired in 8- to 9-year-old congenitally deafened cochlear implant recipients were associated with educational settings that emphasize oral communication development and placement with hearing peers. Less accurate speech intelligibility scores at 8 to 9 years of age were associated with educational programs that emphasize the development of language via signs and placement in special education classes.

So, as a way of summarizing their findings, the study’s own authors are careful not imply that the educational setting was the cause of the intelligibility, but rather that there was an association.

In fact, they discuss this point rather specifically.

One cannot conclude from these results that oral communication mode or mainstream class placement causes children with cochlear implants to develop high levels of speech intelligibility. Children with a propensity for spoken language may be guided toward oral and mainstream programs. Mainstream classroom placement before and immediately after cochlear implantation was not a significant predictor of later speech development. However, mainstreaming became a significant predictor of speech acquisition with increased cochlear implant use. This result suggests that as children became more intelligible, mainstream placement and greater interaction with normal-hearing peers was the result.

In spite of this care not to draw too strong a conclusion, they do still state that “Communication mode, on the other hand, was consistently predictive of speech outcome.” So they’re saying that if a child is in an environment where speech is the only communication mode that their eventual speech intelligibility will be higher. That’s good information actually. And it’s not counter-intuitive. For the children in oral only environments it’s sink or swim. They are under a lot of pressure to articulate correctly because they have no other way to communicate.

Here is how they did the test:

Intelligibility measures were obtained by having the children repeat 36 sentences varying in length from 3 to 5 syllables. Embedded in the sentences were key words selected from a corpus of words known to predict speech intelligibility in deaf children. The speech samples were recorded, edited under computer control, and played to judges who designated by questionnaire that they had had no exposure to deaf speech. All judges signed consent forms approved by the University of Texas at Dallas institutional review board. Judges were allowed to hear a child speak only once and a sentence only once to avoid familiarity effects. Judges wrote the entire sentence. Correct key words were determined and averaged across the 3 judges per sentence per child. Therefore, the average speech intelligibility scores reported were obtained from 108 judges (36 key words times 3 judges) per child.

So it’s an excellent test of pure articulation.

I think it’s significant that we don’t have any comparison data presented in this study. They did not include any similar data for kids with normal hearing. That’s critical to be able to understand what a “normal” distribution looks like. Without any attempt at a control group of any kind this study is not a scientific experiment, it’s a survey. Also, the larger debate includes deaf children who are not cochlear implant candidates for a variety of reasons.

I don’t want to attack the survey or its methods too much, because it’s a nice, tightly focused investigation which is very informative.

But there are many inputs which are not considered or controlled. Further, there is a wealth of ancillary data that was either not gathered or not included in the article. Just because the children can repeat sentences intelligibly, does that mean that they can express themselves well? This study refers to other studies regarding intelligibility and speech development, so we’ll need to look at those studies for answers to the following question: Does having intelligible speech mean that they are using language effectively to communicate with others?

And I still have other questions: Are they learning things other than how to talk? How are their grades? How are their friendships? How do they get along with their families? What happened to these 8 and 9 year-olds later in life? Those are important questions that are not answered by this survey.

The study supports the idea that expectations and environment are important. A child who is expected to speak and whose environment requires it apparently does speak more intelligibly. But does this mean it’s a good idea to avoid signing with all deaf children? Does this mean it’s harmful to sign with deaf children who will get cochlear implants and be expected to use spoken language well?

This study is informative, but it hardly bolsters the case of the oralists outside of a very limited situation. And even if this survey is conclusive on its own, it leaves a number of important questions unanswered. It doesn’t address any broader risks that the oralist approach might carry with it.

Finally, as a conclusion, I like to share a link to a short write-up from Marc Marschark in response to the question, “Is there any evidence that providing or exposure to sign language to cochlear implanted children could hinder speech development?” His answer?

Actually, the research is very clear that sign language does not hinder speech development in children with cochlear implants… in fact it may facilitate it.