What Deaf and Hard of Hearing children need is Language

by Matt Hall, Wyatte Hall and Naomi Caselli

(Note: video is in American Sign Language.)

Note: this is a blog post about a longer article:
Hall, M. L., Hall, W. C., & Caselli, N. K. (2019). Deaf children need language, not (just) speech. First Language, 1-29.

The ability to use language is important for all people. For most people, language acquisition happens naturally and automatically in early childhood. By age 5, most children have basically mastered all major parts of their native language(s), without needing formal instruction or therapy.  This is true for children who have effortless access to language in their everyday environment such as hearing children, and also for Deaf children who are raised by proficient signers. Unfortunately, the situation is different for a majority of deaf or hard of hearing (DHH) children born into families that use primarily spoken languages and do not know a sign language. For them, language acquisition is often delayed and/or incomplete by age 5, which means that these DHH children, in contrast to most other peers, enter school without the language foundation that will be required for success in the classroom and beyond.

Because of this, mastery of at least one language – spoken or signed – needs to be an urgent priority for DHH children. Unfortunately, the current systems that are in place to support DHH children’s development don’t usually frame things this way.  Instead, these systems typically privilege the acquisition of a spoken language. This can happen explicitly (e.g. professionals specifically dissuading parents from using other languages and forms of communication other than spoken language) or implicitly (e.g. parents having a default preference for their child to share their home language, which is almost always spoken). 

In any case, the result is typically that early intervention with DHH usually focuses on the acquisition of a specific language (usually English), rather than at least one language (which would include spoken languages like English and Spanish, but also natural sign languages like American Sign Language – ASL). Note that other forms of manual communication (e.g. sign-supported speech, baby sign, Signed Exact English, and Cued Speech) are all methods of expressing the grammar of English and are invented codes, rather than natural languages. This latter approach of prioritizing acquisition of at least one language reframes the situation from the child’s perspective, asking what kind of input would be easiest for the child to learn. The empirical evidence is fully consistent with the hypothesis that providing DHH children access to a sign language like ASL during their critical language-learning years will help their cognitive, academic, and social-emotional development.

Of course, there are many researchers and other allied professionals who disagree with our view. Next, we identify several of the main objections that these people make, and explain why we do not find them convincing.

Objection 1: “Kids who sign have worse outcomes than kids who don’t sign“

One common objection comes from studies that report an association between increased use of manual communication and decreased performance in proficiency in spoken and/or written language – Geers and colleagues (2017) is a recent example.  They report that DHH children who never used manual communication scored significantly better on a test of spoken English proficiency than DHH children who used manual communication, either short-term or long-term. Based on these results, they recommend that families focus exclusively on listening and spoken language. 

There are three major reasons that we are not convinced by this interpretation. First, they did not show that using sign language was responsible for the poorer performance in spoken language; it’s entirely possible that these children used manual communication because their spoken language skills were not developing, rather than the other way around. This study does not allow us to resolve this chicken-and-egg problem. Second, their study did not differentiate ASL from other forms of manual communication, which – as noted above – are all different forms of English. Therefore, these results never directly tested our hypothesis that early access to ASL would benefit these children. Third, these authors did not consider whether the children did or did not develop proficiency in any language other than English (i.e., ASL). It is possible –although unknown – that the likelihood of achieving age-appropriate language proficiency in at least one language was greater in the groups that used manual communication, since these children might have been able to successfully acquire a sign language. Since these researchers did not assess the children’s signing skills, this remains unknown.  What is known, however, is that even in the best-performing group of children, 49% scored below the 16^th percentile in spoken English in the early years of their elementary school education; for these children, English is the only language they have.  This means that nearly half of the children in the best-performing group lacked age-appropriate skills in any language as they began school. This is a serious delay that sets these children up for a long and difficult struggle inside and outside the classroom. 

It is also worth noting the possibility that the children in the signing groups may not have succeeded in mastering a natural sign language. In this case, it is important to consider what opportunity the child has had to learn such a language. The groups in Geers et al. (2017) and in nearly all other similar studies are based on “communication mode”, but there are serious problems with the way that “communication mode” is defined.  For example, it fails to provide a cumulative history of a child’s input, distinguish between different forms of manual communication, describe the extent of a child’s access to different forms of input, or capture the extent to which a child has lacked access to any kind of input. If we want to find out what kinds of early linguistic input are most likely to yield successful acquisition of at least one language, we will need to move away from “communication mode” and develop better methods of describing the language that children have access to.

Objection 2: “Research shows that sign language prevents kids from learning to hear”

Several different objections fall under this heading. We will briefly explain and address each one.

Their objection: Using sign language takes over the brain’s auditory cortex, making it harder to learn how to listen and speak.

Our rebuttal: There is no causal evidence that this is true – the only evidence available is correlational. There is a major confound in these studies, most of this research only tests children who have lacked access to both sound and language. We can test this hypothesis by examining development in Deaf children with cochlear implants who are raised by fluent signing parents. If this hypothesis is correct, they should have the most difficulty learning to acquire spoken language with their CIs; instead, all studies to date show that these children excel at spoken language acquisition.

Their objection: Hearing is necessary to support cognitive development, especially for executive function and implicit learning.

Our rebuttal: When implicit learning is measured properly, DHH children are not at risk. The problems of executive function are real, but are more likely to be the result of delayed/incomplete access to language. Again, Deaf children who have had access to language from birth do not show any problems in executive function.

Their objection: Decades of research demonstrates that spoken language is necessary for broader developmental success.

Our rebuttal: This view willfully ignores an equally substantial scientific literature demonstrating that proficiency in a natural sign language confers all of the benefits that proficiency in a spoken language does.

Their objection: There is a critical period for spoken language acquisition, so it’s urgent to provide access to spoken language as soon as possible.

Our rebuttal:  The critical period for language acquisition applies to natural sign languages as well. What matters most is to provide the child with access to a type of input that will let the child develop mastery of a language as soon as possible.

Their objection: The earlier we provide auditory access and listening and spoken language therapy, the better the outcomes.

Our rebuttal: This statement is relatively true, but it does not mean that outcomes are good. To date, estimates derived from the peer-reviewed research literature suggest that DHH children whose families emphasize listening and spoken language to the exclusion of natural sign languages typically score between 1 and 2 standard deviations below the mean on tests of spoken language acquisition. This is unacceptable.

Objection 3: “Using sign language might be nice, but it’s just not practical for most families. Be realistic.”

Their objection: Quantity matters – hearing families will be able to provide more input in spoken language than in sign.

Our rebuttal:  What matters most is not how much input a family produces, but rather how much input a child receives. Given that most insurance will not cover cochlear implantation before 12 months, that means that a child without access to a sign language has already missed an entire year of linguistic input – orders of magnitude beyond what classic studies in hearing children focus on.  Most DHH children can access sign earlier than they can access spoken language; using a natural sign language therefore may be the best way to increase the quantity of input that a child has access to. 

Their objection: Quality matters – hearing families will never sign well enough to be appropriate language models for their children.

Our rebuttal: There is likely to be a tradeoff between linguistic quality and perceptual quality.  We know that children are able to regularize inconsistent input, at least when they have perceptual access to it.  It remains entirely possible that hearing parents who are learning ASL can be effective language models for their children, especially in combination with other interventions such as Deaf mentors.

Their objection: The family wants their child to listen and talk, so that’s what we’re focusing on.

Our rebuttal: It is the responsibility of allied professionals to ensure that parents are making an informed choice about their child’s upbringing.  Counseling parents through these conversations must involve discussing the likely risks and benefits of all options, and the likely outcomes of these options.  Currently, options that do not include a natural sign language greatly increase the risk of language deprivation and associated developmental consequences – are parents aware or told of this? These professionals must be very careful in the language they use when counseling parents, so as not to create misleading impressions. 

It’s important to remember that any ambitious goal appears to be impractical at first. This was true for things that are now considered routine, such as universal newborn hearing screening and early cochlear implantation.  We argue that the generally healthy development observed in Deaf native signers provides strong justification for evaluating the extent to which early access to natural sign languages may confer these benefits to Deaf children from hearing families as well.

Now what? Some final recommendations.

We close by providing some recommendations to researchers, parents, and allied professionals. The full list is available in our actual paper. Here, we highlight only a few of what we see as the most important recommendations.

For parents:

• Stay focused on doing everything you can to ensure that your child develops mastery of at least one language – spoken or signed – by the time they’re ready to enter school.  Let this goal guide the other choices you make.
• In your Individualized Family Service Plan, ensure that your child’s progress is being tracked over time, in language and in other domains.  If your child’s gap is getting wider instead of narrower, it’s urgent and necessary to consider different approaches.
• Do not wait for your child to “fail” on one path before beginning to pursue multiple routes to success.

For allied professionals:

• Ensure that families understand that language (not hearing, speaking, signing, or cueing) is the key to unlocking their DHH child’s developmental potential.  If they express a preference for choices that you know to be risky, make sure they fully understand the risks.
• Assess developmental across multiple domains, and assess all languages that are relevant to the child.
• Think critically about the results of various research studies; there may be other interpretations that are equally consistent with the data.

For researchers:

• Stop using “communication mode”, and start describing the child’s cumulative experience with different types of input (or lack thereof).
• Given correlational findings that are compatible with contradictory theories, seek out test cases that can serve to discriminate these contrasting theoretical accounts.
• Work to improve the measurement of language input, language proficiency, and other developmental outcomes in DHH children.

Conclusions

Despite success stories at the individual level, the current understanding of “best practices” has not resulted in satisfactory outcomes at the population level. Large-scale studies continue to find that DHH children on the whole tend to lag far behind their hearing peers in their mastery of a first language, and in many related domains of development – this is language deprivation on a mass scale. This situation calls for the exploration of additional theoretically- and empirically-grounded alternatives that have potential to significantly improve DHH children’s developmental outcomes.  We argue that the developmental benefits of early access to natural sign languages remain largely untested, although all available data lead us to expect beneficial results.  Evaluating whether this is true is therefore an urgent priority for current and future research.

_________________________________________________