Early Sign Language Exposure and Cochlear Implantation Benefits

Even if research such as Geers et al. or other similar studies in the future might provide or appear to provide more definitive evidence that using sign language along with spoken language might cause some initial delays in acquiring either or both languages, well-designed long-term longitudinal studies of deaf children as they grow up into adults are necessary to fully ascertain the overall benefits of using spoken language and sign language with deaf children. Many recent studies have confirmed various benefits of bilingualism, including improving executive functioning in toddlers1 and delaying the onset of dementia in older adults2. While it is possible that teaching deaf children both sign language and spoken language might lead to slight lexical delays2 during their formative years, we should not overlook the long-term benefits of bilingual deaf adults who have stronger executive functioning1,2 and are more likely to be well-adjusted, being able to function and participate in both deaf and hearing communities.3 Therefore, I, as a profoundly deaf-since-birth developmental-behavioral pediatrician, strongly encourage future research of language development in deaf children to adopt a more long-term and holistic approach in understanding and appreciating the value of both spoken and sign languages for all deaf children rather than find ways to negate one or the other approach. From a deaf epistemology perspective3, there is truly no need to conduct and publish short-term research that discourages anybody from learning any language, especially discouraging parents of deaf children from learning sign language while we encourage non-deaf babies to learn sign language. On the other hand, there is a great need for well-designed, scientifically-sound, and community-based participatory research4 to focus on what we can do to improve deaf people’s health and long-term outcomes5.

1. Crivello, C., Kuzyk, O., Rodrigues, M., Friend, M., Zesiger, P., & Poulin-Dubois, D. (2016). The effects of bilingual growth on toddler’s executive function. Journal of Experimental Child Psychology, 141, 121-32. doi: 10.1016/j.jecp.2015.08.004
2. Alladi, S., Bak, T. H., Shailaja, M., Gollahalli, D., Rajan, A., Surampudi, B., Hornberger, M., Duggirala, V., Chaudhuri, J. R., & Kaul, S. (2017). Bilingualism delays the onset of behavioral but not aphasic forms of frontotemporal dementia. Neuropsychologia, 99, 207-212. doi: 10.1016/j.neuropsychologia.2017.03.021
3. Hauser, P. C., O'Hearn, A., McKee, M., Steider, A., & Thew, D. (2010). Deaf epistemology: Deafhood and deafness. American Annals of the Deaf, 154(5), 486-492. doi:10.1353/aad.0.0120
4. Smith, S. R., & Chin, N.P. (2012). Social determinants of health disparities: deaf communities. In J. Maddock (Ed.), Public Health - Social and Behavioral Health, 449-460. Rijeka, Croatia: InTech.

5. Barnett, S., Klein, J.D., Pollard, R. Q., Jr., Samar, V., Schlehofer, D., Starr, M., Sutter, E., Yang, H., & Pearson, T. A. (2011). Community participatory research with deaf sign language users to identify health inequities. American Journal of Public Health, 101(12), 2235-8. doi: 10.2105/AJPH.2011.300247

In response to the numerous criticisms of MS#2016-3489 (Early sign exposure and cochlear implant benefits), Geers et al. state “we investigated whether more focused spoken language exposure, free from distraction of manual signs, would provide a more attainable broad-based strategy for verbal language development.” Further, they explicitly cite my work to support the conclusion that, because of differences in the structure of ASL and English, “. . . conscientious and proficient use of ASL would detract from the amount of time spent stimulating and reinforcing spoken language development and could influence the structure of spoken language.” Nothing in the cited work remotely supports such a conclusion - in neither my 2002 book “Language, Cognition, and the Brain: Insights from Sign Language Research” nor in the 2005 article on code-blending by hearing adult ASL-English bilinguals that Geers et al. cite. On the contrary, recent research clearly demonstrates that simultaneously perceiving signs and spoken words does not negatively impact spoken word recognition or learning in deaf children1 and facilitates word recognition in adults2. Further, the augmentative use of signs has recently been shown to be beneficial to word learning in deaf children3.

In addition to mis-citing the literature, the response does not address the “correlation is causation” fallacy brought up by several commentators. In the article, as well as in their response to comments, the authors strongly imply that long-term use of ASL with deaf children causes poor English outcomes, for example because use of sign language is distracting or because parents spend less time focused on English. The latter concern is one that is commonly voiced by parents who are raising hearing children with two spoken languages. However, numerous studies have shown that children who are exposed to two languages are not impaired in language learning compared to children exposed to a single language. This is despite the fact that they receive less input in a given language since their parents divide their language use between two languages. Nonetheless, hearing parents are learning ASL, and more research is needed on how the quality of parental input impacts linguistic development is such bilingual settings.

The methodological concerns about data analysis and interpretation remain. The experimental design still suffers from probable self-selection biases – as Geers et al. acknowledge in their response, parents may choose to continue to use sign because they struggle to communicate with their child using spoken language alone. The fact that ASL was not distinguished from artificial sign systems (e.g., Signing Exact English, Total Communication) remains a fatal flaw because of known problems with the use of these systems with deaf children4, and this means that the authors cannot draw any conclusions about the impact of sign language on spoken language development. Overall, the interpretation that use of sign language interferes with auditory and speech development is simply not warranted, and this conclusion is dangerous given the harmful effects of language deprivation to the cognitive and social well-being of deaf children. Nothing in Geers et al.’s response changes this fact.

1. Giezen MR, Baker AE, Escudero P. Relationships between spoken word and sign processing in children with cochlear implants. J Deaf Stud Deaf Educ. 2013;19(1): 107-125.

2. Emmorey K, Petrich JAF, Gollan TH. Bilingual processing of ASL-English code-blends: The consequences of accessing two lexical representations simultaneously. J Mem Lang. 2012; 67: 199-210.

3. van Berkel-van Hoof L, Hermans D, Knoors H, Verhoeven L. Benefits of augmentative signs in word learning: Evidence from children who are deaf/hard of hearing and children with specific language impairment. Res Dev Disabil. 2016; 59: 338-350.

4. Strong M, Charlson ES. Simultaneous communication: Are teachers attempting an impossible task? Am Ann Deaf. 1987; 132(5): 376-382.

Response to Comments on “Early Sign Language Exposure and Cochlear Implant Benefits”

By:  Ann E. Geers^a, Ph.D, Christine M. Mitchell^b, ScM, Andrea Warner-Czyz^a, Ph.D., Nae-Yu Wang^b, Ph.D., Laurie S. Eisenberg^c, Ph.D.

^a Callier Center for Communication Disorders, University of Texas at Dallas

^b Welch Center for Prevention, Epidemiology and Clinical Research, Johns Hopkins University

^c Keck School of Medicine of the University of Southern California, Los Angeles

--------

1. Issues with Regard to Design of the Study

Sound scientific research evaluating the effects of an intervention is predicated on random selection or assignment to treatment and control groups. The Geers et al. (2017) study violates this basic tenet of research design

While we agree that randomly assigning families at the time of diagnosis to intervention programs that differ in their use of sign language would be an efficient method to address sign language benefits, this approach is impractical and unethical for a variety of reasons.  A decision regarding a child’s communication mode arises from parental choice based on information available at the time the child becomes a candidate for cochlear implantation.  The current observational cohort study examined more prospectively collected pre-implant demographic characteristics than any previous study on this topic in the literature, introducing unprecedented transparency into the interpretation of group comparability at baseline.

2. Issues with Defining Exposure to Sign Language

…although the authors quantified exposure to manual communication (including ASL), they did not measure proficiency in ASL. Children who have not acquired the grammar of ASL are not predicted to benefit from it as a foundation for subsequent mastery of English….

Effects of ‘sign language exposure’ may have been carried by participants who used an artificial signing system, received late exposure relative to the critical period of language acquisition, had only one ASL model, and families with limited-to-no ASL proficiency….

From study onset, parental and clinician goals for participants in the CDaCI study have focused on spoken language development, and that outcome drove the research questions in this paper.  We sought to examine long-term consequences of early communication mode decisions made for deaf children receiving cochlear implants (CI) by parents who have typical hearing.  

Our study includes data from a representative sample of early-implanted children reared in real-world situations across the United States, in which the majority of hearing parents typically lack proficiency in American Sign Language (ASL), and, therefore, cannot provide a language-rich environment in both ASL and spoken English.  Differences in syntax, phonology, and spatial information between ASL and spoken language^1,2 limit simultaneous use of both languages. Therefore, conscientious and proficient use of ASL would detract from the amount of time spent stimulating and reinforcing spoken language development and could influence the structure of spoken language. Therefore, most hearing parents choose to employ an “artificial” sign language to accompany spoken English. 

This paper does not aim to examine the effects of proficient ASL exposure on outcomes in pediatric CI users.  Rather, we investigated whether more focused spoken language exposure, free from distraction of manual signs, would provide a more attainable broad-based strategy for verbal language development.

3. Issues with Pre-Existing Group Differences

… the best interpretation of this study is that families self-select their method of communication as a result of their child’s development in English. Under this view, the use of manual communication is a consequence of limited progress in English, not a cause.

Parents of deaf children who are not progressing with their CI may be more likely to begin, or continue, signing with their child. This would imply that poor oral outcomes encourage use of signing, rather than the use of signing limiting oral outcomes.

Baseline measures (Table 1) and communication mode decisions were made in infancy (i.e., before the child was 38 months old). Children in the three groups exhibited similar spoken language at that time (actually children in the long-term sign exposure group had the highest average spoken vocabulary size pre-implant), making “self-selection” to use sign on the basis of limited progress in spoken English questionable. Average baseline auditory perception ratings do not differ significantly among groups. Pre-implant aided pure-tone average thresholds differed by less than 5 dB across groups. Average parent ratings of auditory-based behaviors on the Infant-Toddler Meaningful Auditory Integration Scale (IT-MAIS) represent very limited auditory skills with hearing aids prior to implantation in all groups – mean scores of 9.8 and 5.8 in the no sign group and the long-term sign groups, respectively – out of a maximum score of 40.

These data indicate the sign exposure groups were experiencing profound difficulty with auditory perception and speech identification. It will surprise no-one that children with better auditory perception skills and speech identification skills at the outset will be the children who end up with the best auditory language scores.

The three participant groups did not statistically significantly differ on auditory perception or spoken language skills at study onset. Furthermore, baseline auditory perception scores were not significantly correlated with any outcome measures (r values ranged from -.02 to .19).  Table 2 shows emerging differences over time in speech perception abilities, measured via an index based on a hierarchy of tests. Two possible interpretations can explain the widening of the gap between performance of the no-sign and sign-exposed groups over time:  (a) Parents of children who do not receive adequate auditory benefit from their CI continue or begin manual communication efforts to interact more efficiently with their child; or (b) Children reared in families using sign develop auditory perception skills more slowly than those reared in non-signing families.  Our results appear to support the second explanation, as all three groups of children exhibited improved speech perception scores over time.  Signing parents typically initiated sign language use before their child received a cochlear implant and none of the no-sign families began to sign later, in response to poor auditory or spoken English progress. Parents in the long-term sign language group reportedly decreased their use of sign over time (from 63% of the day at baseline to 29% at 36 months post-implant). The 27 families in the short-term sign group stopped using sign during the first two years of CI use, presumably because it was no longer necessary for communication, though their outcomes did not reach the high levels achieved by the no-sign group.  Nevertheless, we cannot state unequivocally whether parents’ use of sign was in response to their child’s auditory development or whether children’s auditory speech perception was negatively affected by parental sign use.

. ….observed differences might be attributable to other demographic factors that likely affected initial inclusion, attrition over time, and/or performance on the assessments (e.g. SES, etiology of deafness). Because sign language use may co-vary with these additional factors, the reported effects might well disappear if these factors were controlled.  To satisfactorily demonstrate that sign language exposure harms spoken language development, the authors must demonstrate that: 1) all baseline measures were equivalent, 2) groups were not self-selected, and 3) participant attrition was not systematic.

It is important to recognize the large variability in all of these characteristics, which explains why differences in average scores cannot be interpreted without recognizing the substantial overlap in distributions.  While statistical analysis failed to indicate significant differences in baseline metrics, some averages favor one group and some averages favor another group (e.g., higher maternal education and family income appear to favor the no-sign exposure group, while the long-term sign exposure group exhibits later onset of profound deafness, higher average IQ, and greater exposure to parent-infant intervention prior to cochlear implantation).  Correcting for all of these variables in the analysis with multiple outcomes and time-points would have added an unnecessary level of complexity to the manuscript, so presumed equivalence was based on failure to find statistically significant group differences.  Furthermore, some pre-implant characteristics (e.g, auditory speech perception) were measured with rating scales that are, at best, ordinal in nature, limiting their usefulness in correction analyses.

However, as a result of receiving several different comments related to potential for baseline group differences, though not statistically significant, to be driving the statistically significant group differences in later outcomes, we have run regression models adjusting for implant activation age, IT-MAIS score, enrollment in a parent-infant program, maternal education, and household income at baseline. It should be noted that, due to substantial correlations among some variables, reported predictors are subject to suppression effects.  Results of these regression models are presented here in Table A.  Even after adjusting for the mentioned characteristics, the groups exposed to sign language performed significantly poorer than participants not exposed to sign language on spoken language (CASL) assessed in the late elementary years and on speech intelligibility.

Table A.  Adjusted linear regression results comparing the sign exposure groups on language, reading, and speech intelligibility outcomes.

* < 0.05, ** < 0.01, *** < 0.001

The 40 children who met eligibility criteria but were excluded due to lack of follow-up data may have influenced the outcomes. Families experiencing poor progress with their child’s CI may stop their follow-up appointments, for instance.

As was stated in the article, 40 children who met the implant age criteria were excluded due to lack of sufficient follow-up data, many of whom moved away from the implant center and were unable to return for follow-up visits.  Similar rates of sign language use at baseline were reported for these 40 children (55% sign - 45% no sign) as for the sample with complete data (60% sign, 40% no sign).  There was no significant tendency for participant attrition to be associated with either signing or non-signing families pre-implant. 

4. Issues with Regard to Interpreting Outcomes

Although this study was designed to look narrowly at English-based outcomes, the authors over-interpret the results as evidence against the assertion that a natural sign language can be beneficial for deaf children. While English proficiency is certainly one route to success, it is not a necessary condition for it.

Our study should not be considered an indictment on the use of ASL, but rather it adds new information to the literature on this age-old controversy.  One of the most consistent findings of CDaCI has been the large variability in outcomes that has persisted throughout the 15 years of data collection. The many reports from this study have examined this variability without focusing exclusively on the “star performers”. In fact, one recent study by Barnard et al reported specifically on those children who had not progressed in their auditory skill development (i.e., did not achieve open-set speech recognition) after five years of experience with the cochlear implant.³ Without adequate access to speech, visual language becomes essential whether it be speechreading, cued speech, simultaneous communication, or ASL. This was the case for the children in the Barnard et al. study. As has been mentioned earlier in this response, the overarching aim of CDaCI has been to study the development of spoken language following cochlear implantation, which injects a certain bias into the research design.  Each family needs to make their own informed decision about what works best for their unique situation.  One of the most important variables in any parent-child relationship, regardless of auditory status, is the presence of effective and efficient communication.  Parents need to keep open lines of communication between parent and child and to provide a language-rich environment to enhance that child’s listening and language development.

REFERENCES

1. Emmorey, K. (2002). Language, cognition, and the brain: Insights from sign language research. Mahway, New Jersey: Lawrence Erlbaum Associates.

2. Emmorey, K, Borinstein, H, Thompson, R (2005) ISB4: Proceedings of the 4th International  Symposium on Bilingualism, ed. James Cohen, Kara T. McAlister, Kellie Rolstad, and Jeff MacSwan, 663-673.  Somerville, MA: Cascadilla Press.

3. Barnard, J.M., Fisher, L.M., Johnson, K.C., Eisenberg, L.S., Wang, N-Y., Quittner, A.L., Carson, C.M., Niparko, J.K., and the CDaCI Investigative Team. (2015). A prospective, longitudinal study of US children unable to achieve open-set speech recognition five years after cochlear implantation. Otology & Neurotology, 36, 985-992.

The utility of sign language as a supplement to spoken language for children using cochlear implants is complicated to systematically study. This topic has been debated for many years and yet there remains a paucity of high quality data to help parents make decisions that may affect their child and family. The Geers et al article is an effort to further our understanding of this topic of interest. The data comes from a large and well-respected database, the methodology and analysis have satisfied the peer review process and thus the findings have appropriately been published. Like most good research, it raises many more excellent questions for future study and has generated discussion. This is a healthy and necessary process for objectively advancing our knowledge.

Children born deaf are a heterogenic population. This prospective study by Geers et al provides insight into a distinct population of children with cochlear implants. Specifically, the children studied were implanted at a young age, had no other disabilities, were managed at centers with specialized personnel, and had highly educated mothers. The children and their families received expert spoken language therapy, with or without varied amounts and types of sign language. In this population, supplemental sign language did not provide an advantage for the development of spoken language. A study using consistent approaches for sign language and randomization might provide more definitive insight but doing such a study remains impractical and ethically challenging.

The Geers et al findings published in Pediatrics may be interpreted as threatening to the values of some individuals. Indeed, interpretations that draw a causal relationship between sign language usage and poor spoken language development would be inappropriate. Likewise, discounting the findings of this study, which by all accounts is the highest quality study carried out to date on this topic, would be similarly unfitting. We are left to balance the findings of this study with the previous literature and our clinical acumen to make recommendations for patients and their families. This study is yet another piece of useful information in the complex puzzle those families must solve as they seek the best opportunities for their children’s futures. As clinicians, we should continue to personalize language recommendations for each child who is deaf, based on their individual circumstances.

Colin Driscoll MD
Chair, Board of Directors, American Cochlear Implant Alliance

A goal shared by all parents of congenitally deaf children is for their child to achieve language competencies that provide a foundation for nurturing, education, and engagement in society. Today, even against a backdrop of miraculous technological advances and an increasing sophistication in our understanding of human capacity for language acquisition, we fall short of being able to provide a prescription for individual deaf children which will guarantee spoken language competencies. In the absence of a definitive solution we must rely upon rigorous research to make correct inferences in ascribing best practices for parents with deaf children.
Sound scientific research evaluating the effects of an intervention is predicated on random selection or assignment to treatment and control groups. This ensures that each participant or subject has an equal chance of being placed in any group. Random assignment of participants helps to ensure that any differences between and within the groups are not systematic at the outset of the experiment. Thus, any differences between groups recorded at the end of the experiment can be more confidently attributed to the experimental procedures or treatment.
The Geers et al. (2017) study violates this basic tenet of research design. The incorrect assumption made by Geers et al. (2017) is that all of their subjects have an equal potential to develop auditory and spoken language skills, or more accurately that this potential will be normally distributed across the No-sign, Short-term Sign, and Long-term Sign groups. However, the authors grouped the children by their experiences with spoken language and manual communication after the parents and children had already chosen a communication strategy. The reality is that deaf children who are not achieving gains in speech and spoken language skills are recommended to augment spoken language skills with some form of manual communication. The selection bias evident undermines the subsequent comparisons made across these groups.
Baseline data presented in Tables 1 and 2 support this assertion. On the Auditory Perception test the Long-Term Sign group received the lowest scores (5.8) followed by the Short term Sign-Group (7.0) with the No Sign Group showing the best scores (9.8). On the Speech Recognition in Quiet (SRI-Q) prior to cochlear implantation, the Long Term Sign group had the lowest and least variable scores (7.5; 1.3-18.8) compared to the No Sign Group (16.3; 5.0-50.0) who showed the highest scores with the greatest variability, the Short-term sign group was intermediate (15.9, 2.5-22.5). These data indicate the sign exposure groups were experiencing profound difficulty with auditory perception and speech identification. It will surprise no-one that children with better auditory perception skills and speech identification skills at the outset will be the children who end up with the best auditory language scores.
Researchers and editors alike must be vigilant in putting forth rigorous research that address the factors that limit language potential in deaf children with cochlear implants. Publishing research with inherently flawed research designs is a disservice to all parents who want their deaf children to flourish.

Regarding Geers et al.’s methodology, data analysis, and conclusions, we find general agreement with other comments, but here we focus on their underlying epistemology. Geer et al.’s epistemological approach focused on understanding how to help deaf and hard of hearing (DHH) children function within a hearing world. The scientific method, however, requires an understanding that science is not neutral – as has been noted over the past centuries with regards to other minority groups. If one accepts the goal of raising DHH children to pass as hearing (1), then the main focus on spoken language acquisition fits with the epistemology of a DHH child as having a disability. The disability epistemology that Geers et al. utilizes, however, accepts the high probability of curtailing and restricting the cognitive and linguistic potential of DHH children.

Cognitive ecology recognizes the importance of embodied cognition, meaning that one’s environment shapes or limits one’s cognition (2). These holistic views of a DHH child as one who is different – rather than one with a deficit – is supported by a Deaf epistemology (3). This perspective facilitates a supportive cognitive ecology with the deaf individual as a visual being. Given this difference, most – if not all – DHH individuals will benefit from an alignment of their cognitive ecology with their own sensory capabilities provided by Deaf epistemology.

From a Deaf epistemology framework, it is worthwhile to note references not included in Geers et al.’s introduction. In particular, the absence of the work of the Early Educational Language Study – which found that beginning signing hearing parents’ children had better pre-literacy skills than those children whose parents did not sign – and studies showing that bimodal bilinguals have better academic outcomes (4) is especially striking. Another recent study (5) found that regardless of having or not having a cochlear implant, or preferring sign or spoken language, there were few differences between high school and college DHH students in their executive functioning and social emotional development. A comment from this study is worthy of additional research in consideration that most of their participants had both sign and spoken language by the age of three: Does this finding suggest that Deaf Education is changing?

With the newer hearing technologies, DHH children are frequently seen as bimodal bilinguals (4). Many DHH students are proud of their ability to live in both the Deaf and hearing worlds, and these young people are more and more common. The evidence from peer reviewed literature and other comments directed against Geers et al., therefore, demand that we all broaden our perspectives and evaluate our epistemologies, and learn to collaborate in elucidating factors that permit and prevent DHH children’s ability to thrive.

References
1. Harmon, K. (2013). Growing up to become hearing: Dreams of ‘passing’ in oral deaf education. In J. Brune & D. Wilson (Eds.), Disability and passing: Blurring the lines of identity (pp. 167-198). Philadelphia, PA: Temple University Press.
2. Hutchins, E. (2010). Cognitive ecology. Topics in Cognitive Science, 2(4), 705-715. doi:10.1111/j.1756-8765.2010.01089.x
3. Hauser, P. C., O'Hearn, A., McKee, M., Steider, A., & Thew, D. (2010). Deaf epistemology: Deafhood and deafness. American Annals of the Deaf, 154(5), 486-492. doi:10.1353/aad.0.0120
4. Leigh, I., & Andrews, J. (2017). Deafpeople in society: Psychological, sociological, and educational perspectives (2nd ed.). New York, NY: Routledge.
5. Marschark, M., Kronenberger, W. G., Rosica, M., Borgna, G., Convertino, C., Durkin, A., ... & Schmitz, K. L. (2017). Social maturity and executive function among deaf learners. Journal of Deaf Studies and Deaf Education, 22(1), 22-34. doi:https://doi.org/10.1093/deafed/enw057

The data presented by Geers et al. are consistent with three mutually-exclusive theories about the impact of natural sign languages on the development of English language skills: that natural sign languages (1) impede, (2) facilitate, or (3) have no impact on English development. Although Geers et al. clearly favor Theory 1, it is neither the only nor the best interpretation of their data.

Theory 2 is consistent with the data because although the authors quantified exposure to manual communication (including ASL), they did not measure proficiency in ASL. Children who have not acquired the grammar of ASL are not predicted to benefit from it as a foundation for subsequent mastery of English. We applaud these authors for considering variation in the amount of exposure to manual communication; however, we are dismayed to see ASL lumped together with other types of manual communication. Such coarse grouping prevents this crucial hypothesis from being adequately tested. (That is, children who are exposed mainly to English-based signing systems will not acquire the grammar of ASL; their performance is therefore uninformative about Theory 2.) Contrary to their claims, the authors have not falsified the theory that mastering the grammar of a sign language helps a child master the grammar of a spoken language.

Theory 3 is consistent with the data because observed differences might be attributable to other demographic factors that likely affected initial inclusion, attrition over time, and/or performance on the assessments (e.g. SES, etiology of deafness). Because sign language use may co-vary with these additional factors, the reported effects might well disappear if these factors were controlled. Theory 3 therefore remains viable.

In our view, the best interpretation of this study is that families self-select their method of communication as a result of their child’s development in English. Under this view, the use of manual communication is a consequence of limited progress in English, not a cause. Both of these interpretations remain available because the study used a correlational design that was particularly vulnerable to self-selection.

To successfully discriminate among the three competing theories, future research will need to: (1) distinguish ASL from other forms of manual communication, (2) assess ASL proficiency as a function of ASL exposure, (3) adopt a research design that minimizes the potential impact of reverse causality and self-selection effects (for example, studying the impact of exposure during a pre-specified time window on outcomes after that time window, regardless of the family's communication choices at time of test).

Finally, it is important to remember that English language skills are only one aspect of child development. Mastery of any natural language, signed or spoken, is expected to support healthy cognitive and psychosocial development, thereby promoting school readiness and long-term success. We must not forget that even in the group of best-performing children in this study, 49% fell below the average range on language proficiency: in hearing children, this figure would be a mere 16%. Clearly, much more work remains to be done to maximize deaf children’s developmental potential.

We outline a number of fundamental issues in how sign language exposure and proficiency were operationalized and reported by Geers et al. Most importantly, the authors did not distinguish between those exposed to American Sign Language (ASL) versus English signing systems (e.g., signing exact English, sign-supported English, baby sign) when classifying children. This is a fatal flaw because, in contrast to artificial English signing systems, natural sign languages such as ASL are legitimate languages – as long-affirmed by the Linguistic Society of America[1] – with all the cognitive benefits a natural language provides. The study is recklessly misleading because of this inappropriate conflation, especially given that the authors’ conclusions contribute to long-standing bias, resistance, and misperceptions against natural sign languages in clinical recommendations for deaf children.

Among other issues, there is not enough information provided about participants’ sign language proficiency and exposure. At minimum, it is critical to know the number of children exposed to only ASL (as opposed to artificial signing systems), the age of first exposure to ASL, the number of ASL language models, and the ASL proficiency of parents and children. Effects of ‘sign language exposure’ may have been carried by participants who used an artificial signing system, received late exposure relative to the critical period of language acquisition, had only one ASL model, and families with limited-to-no ASL proficiency. The little information provided about sign language exposure was not collected using direct measurement; rather, it appears to have been measured using an unvalidated parental-report questionnaire. The criterion for positive indication of sign language exposure was – in our view – very low (> 10% of the time), and there was no rationale offered for why 10% is minimally-sufficient. It is possible that the sample in this study represents a straw man hypothesis; no one would argue that such language conditions are sufficient for a child to thrive.

ASL is typically used within a bilingual approach encouraging both natural sign language and spoken/written English acquisition[2], and should be evaluated as such. Because those children are emerging bilinguals, their combined proficiency in both ASL and English must be considered to draw any conclusions about ASL-based intervention efficacy. Further, because bilingual and monolingual language acquisition differs, bilingual signing children’s appropriate comparison group are other bilingual children and should not be compared to monolingual norms.

Although this study was designed to look narrowly at English-based outcomes, the authors over-interpret the results as evidence against the assertion[3] that a natural sign language can be beneficial for deaf children. While English proficiency is certainly one route to success, it is not a necessary condition for it. The results of this study have no bearing on whether exposure to a natural sign language has any effect on the holistic well-being and health-related outcomes of deaf children, but they are dangerously framed and misinterpreted as such.

References
[1] Resolution: Sign Languages. Linguistic Society of America. 2001. Available at: https://www.linguisticsociety.org/resource/resolution-sign-languages. Accessed June 18, 2017.

[2] Lange C, Lane-Outlaw S, Lange W, Sherwood D. American Sign Language/English Bilingual Model: A Longitudinal Study of Academic Growth. Journal of Deaf Studies and Deaf Education. 2013;18(4):532-544. doi:10.1093/deafed/ent027.

[3] Napoli DJ, Mellon NK, Niparko JK, et al. Should all deaf children learn sign language? Pediatrics. 2015;136(1):170–176. pmid:26077481.

Based upon an analysis of data collected from the CDaCI study, Geers et al. argue that parents’ use of sign-based communication systems may result in poorer listening and spoken language skills, and reading outcomes for deaf children who have received a CI. Their findings conflict with those reported by Niparko et al. (1), which used the same dataset, raising concerns about variable selection and selective reporting of analyses. Whereas Niparko et al. concluded that signing had no effect on spoken language outcomes as measured using the RDLS, Geers et al. argue that the CASL revealed poorer spoken language and listening outcomes for deaf children whose families have chosen visual communication strategies.

Additionally, they point out that those same children show long-term deficiencies in reading skills as measured by the WJ-IV. Given the large, multivariate CDaCI dataset and the need for robustness in the findings across a range of predictor and outcome measures, we argue that these differences may reflect measurement error. Two data points do not allow for determination of developmental change (2), therefore the data cannot be classified as truly longitudinal. In addition, the data are presented in a grouped format, so actual change for individual children is obscured. The grouped format furthermore exhibits high variability which occludes some of the conclusions made by Geers et al.

Finally, Geers et al report no difference in outcomes for children in families using sign 10-50% of the day versus those who reported using sign >50% of the day. Given no apparent dose-response relationship between sign use and outcomes, this argues in favor of a “third variable” being responsible for both continued use of sign language and the poorer outcome measures in the dataset. For example, failure to obtain significant auditory benefit from the cochlear implant may have led to difficulties in spoken language communication and a family decision to continue to use sign with their deaf child. Socioeconomic factors (i.e. upper-middle class, white, educated) are often linked to better language environment for the child. (3) Households in the signing groups in Geers et al’s study had lower maternal education and income levels (Table 1), and likelier poorer insurance coverage. (4) These factors can contribute to weaker compliance with rehabilitation protocols and thus poorer outcomes. (5)

Because Geers et al. did not distinguish between the different kinds of visual communication parents used with their children, nor were there measurements of visual communication proficiencies in the CDaCI dataset, it is difficult to determine to what extent poor visual communication abilities contributed to lower spoken language abilities in signing families. Overall, the data analysis provide no evidence for causation. It merely hints at a correlation between parental communication choice, latent variables either not measured or not included in the analysis, and selected listening and spoken language outcomes. We agree with Geers et al. that more data is needed to address the questions posed in their study. However, they have not provided suitable answers to them.

References

1 Niparko JK, Tobey EA, Thal DJ, Eisenberg LS, Wang N-Y, Quittner AL, Fink NE, CDaCI Investigative Team. Spoken language development in children following cochlear implantation. JAMA. 2010;303(15): 1498-1506. doi: 10.1001/jama.2010.451.

2 Singer JD, Willett JB. Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. New York (NY): Oxford University Press; 2003.

3 Hoff E. The specificity of environmental influence: socioeconomic status affects early vocabulary development via maternal speech. Child Dev. 2003; 74(5): 1368-1378. doi: 10.1111/1467-8624.00612.

4 Cochlear Implants FAQ. U.S. Food and Drug Administration. 2014. Available at: https://www.fda.gov/MedicalDevices/ProductsandMedicalProcedures/Implants.... Accessed June 18, 2017.

5 Chang DT, Ko AB, Murray GS, Arnold JE, Megerian CA . Lack of financial barriers to pediatric cochlear implantation: impact of socioeconomic status on access and outcomes. Arch Otolaryngol Head Neck Surg. 2010;136(7): 648-657. doi: 10.1001/archoto.2010.90.

Geers et al. contains significant methodological issues that should moderate any findings and claims of sign language’s role in implanted deaf children’s spoken and written English development.

First, the study sample is (understandably) non-randomized; thus categorization factors may be related to outcomes. Asserting a causal conclusion from a correlational non-randomized study is inappropriate – especially when a simpler explanation may exist: parents of deaf children who are not progressing with their CI may be more likely to begin, or continue, signing with their child. This would imply that poor oral outcomes encourage use of signing, rather than the use of signing limiting oral outcomes.

Secondly, while the authors reported no statistically significant differences between groups at baseline, the actual data suggest clinically significant differences that were statistically non-significant due to small group sizes. Multi-layered and complex variables such as maternal education (69% vs 50%), income  <$50k (32% vs 43%) and age of onset (0.3m v 1.2m) are well-known to influence language and reading outcomes (it is also unclear if age of onset is actually age of diagnosis). Additionally, auditory perception abilities at baseline were much lower in the group that continued to sign; indeed, the authors recognize early speech recognition predicts later speech intelligibility. Furthermore, type and frequency of post-implant rehabilitation – an educational experience independent of actual CI benefits – was unaccounted for.

Thirdly, it is unclear how the authors characterized “signing” and “percent of time signing”, or whether parents understood how ASL differs from other signing systems. Moreover, parents may have differed widely in interpreting and estimating the “percent” of time using sign at home. Hence this measurement may not reflect actual amount of signing and may not constitute a valid measurement of sign language exposure.

Finally, the suggestion that using sign language interferes with English language development for all deaf children requires acknowledging critical limitations of subject selection that were not discussed. As with other CI studies, subject selection was biased towards including children who succeed with their CI. The 40 children who met eligibility criteria but were excluded due to lack of follow-up data may have influenced the outcomes. Families experiencing poor progress with their child’s CI may stop their follow-up appointments, for instance. Or, families may decide to stop continuing with the CI and focus on sign language only. Since race and maternal education differed significantly between selected and non-selected groups, baseline data on the excluded families should be reported and evaluated for any “dropout” associations from the study. Additionally, some excluded families may comprise a fourth unreported group: families who did not sign at baseline but began signing during the follow-up periods. The absence of this group is particularly striking.

To satisfactorily demonstrate that sign language exposure harms spoken language development, the authors must demonstrate that: 1) all baseline measures were equivalent, 2) groups were not self-selected, and 3) participant attrition was not systematic. This study design met none of these conditions; we thus find the authors’ conclusions unconvincing at best.