2.1 Sign language linguistics

Research into the linguistic structure of sign languages is still quite young. The most extensive early work emerged primarily in response to American Sign Language (ASL) and British Sign Language (BSL) between the 1960s and the 1980s (Stokoe Reference Stokoe1960; Klima & Bellugi Reference Klima and Bellugi1979; Kyle & Woll Reference Kyle and Woll1985). Since then, interest has developed globally in different sign languages, particularly regarding their role in education. The linguistic structure of sign language has been fully described elsewhere (see, for example, Sutton-Spence & Woll Reference Sutton-Spence and Woll1999) therefore it is sufficient to underline that research has demonstrated that sign languages, as natural languages, fulfil the same linguistic, social and cognitive functions as spoken languages (Emmorey Reference Emmorey2002). A growing literature on sign languages continues to develop our understanding of their linguistic structure (Yang & Fischer Reference Yang and Fischer2002; Hunger Reference Hunger2006) and the social and cultural contexts of their use (Branson & Miller Reference Branson and Miller2004; Slegers Reference Slegers2010; McKee & McKee Reference McKee and McKee2011; Behares, Brovetto & Peluso Crespi Reference Behares, Brovetto and Peluso Crespi2012; De Quadros Reference De Quadros2012; Geraci Reference Geraci2012; Parisot & Rinfret Reference Parisot and Rinfret2012; Quer Reference Quer2012; Schermer Reference Schermer2012).

The visual-gestural characteristics of sign language have prompted extensive interest in the neurological processing of the sensory motor aspects of sign language. The most significant finding for this review demonstrates that the neural systems that support sign and spoken language similarly involve the left-lateralised network of the brain (Hickok, Love-Geffen & Klima Reference Hickok, Love-Geffen and Klima2002). Although there is some right hemisphere activity in sign language processing for spatial and movement encoding, the left hemisphere of the brain shows dominance for sign language processing (Mineiro et al. Reference Mineiro, Nunes, Moita, Silva, Castro-Caldas, Marschark, Tang and Knoors2014). Interestingly, as will emerge in the consideration of literacy development (Section 3), the phonological processing of signs and words is identical (MacSweeney et al. Reference MacSweeney, Capek, Campbell and Woll2008).

2.2 Sign language acquisition

Sign language acquisition studies have focused on how individuals master aspects of sign language structure such as phonology (Morgan Reference Morgan2006; Mann et al. Reference Mann, Marshall, Mason and Morgan2010), spatial language (Schembri, Jones & Burnham Reference Schembri, Jones and Burnham2005; Morgan et al. Reference Morgan, Herman, Barriere and Woll2008), narrative skills (Robertson et al. Reference Robertson, Quintela, Ramírez and Lissi2012) and grammar features (Thompson, Emmorey & Kluender Reference Thompson, Emmorey and Kluender2009). This area of research has been extended to include the study of impairments or disorders in sign language use (Meronen & Ahonen Reference Meronen and Ahonen2008; Mason et al. Reference Mason, Rowley, Marshall, Atkinson, Herman, Woll and Morgan2010; Quinto-Pozos, Forber-Pratt & Singleton Reference Quinto-Pozos, Forber-Pratt and Singleton2011; Woll & Morgan Reference Woll and Morgan2012). Nonetheless, this work is still very much in its infancy and, for many countries, the indigenous sign language has not yet been fully described.

Studies of sign language acquisition in children show that sign language development is comparable to spoken language development. This knowledge is mainly based on research in the USA and the UK with deaf children of deaf parents, demonstrating that early fluency in sign language leads to equivalent word and sentence level milestones and pragmatic skills (Petitto et al. Reference Petitto, Katerelos, Levy, Gauna, Tétreault and Ferraro2001; Most Reference Most2003; Rinaldi & Caselli Reference Rinaldi and Caselli2009). However, this is not the case for the majority of deaf children who are non-native or non-fluent signers. These are usually children of hearing parents who have no prior experience of deafness or sign language. The development of sign language fluency for this group of deaf children is problematic as they experience a general delay in sign language development and difficulties in catching up. This has been evidenced through scores on standardised sign language assessments (Herman & Roy Reference Herman and Roy2006) as well as general language comprehension (Rodríguez Reference Rodríguez2007) and early pragmatic and narrative competence (Becker Reference Becker2009; Surian, Tedoldi & Siegal Reference Surian, Tedoldi and Siegal2010).

This delay is explained in part by the limited access to everyday conversation and routine interactions, such as storytelling or parental commentary around a shared activity. Hearing parents reportedly find it difficult to learn sign language as adults, as a second language (L2) (Napier, Leigh & Nann Reference Napier, Leigh and Nann2007; Von Pein & Altarriba Reference Von Pein and Altarriba2011). Even though many of these families do develop their own ‘homesign’, these communication strategies do not fulfil the experience of early access to a fluent language from birth (Janjua, Woll & Kyle Reference Janjua, Woll and Kyle2002; Morford & Hänel-Faulhaber Reference Morford and Hänel-Faulhaber2011).

The case of deaf children who are born without full access to a spoken language and who have limited or no exposure to a sign language provides insight into the impact of delayed exposure to a first language (either spoken or signed). There are of course many mediating factors which alter individual circumstances (age of diagnosis, aetiology, comorbidity, intervention experience, parental hearing status, access to hearing technology) and which can facilitate the development of a fluent spoken and/or sign language. However, when considering what happens when accessible exposure to sign language as a first language (L1) is delayed, research has focused on neurological development, the overall ability to learn language (Cormier et al. Reference Cormier, Schembri, Vinson and Orfanidou2012), and stages of development (Berk & Lillo-Martin Reference Berk and Lillo-Martin2012). These studies find there are processing deficiencies when access to either a spoken or a sign language as L1 is delayed and that this has a subsequent impact on the development of literacy (Mayberry Reference Mayberry2007; Malaia & Wilbur Reference Malaia and Wilbur2010; Mayberry et al. 2011a).

2.3 Sign language assessment

The growth of understanding of sign language acquisition has led to the development of standardised assessments for sign language in several developed counties and these are also emerging in less developed countries (see, for example, Lichtig et al. Reference Lichtig, Couto, Mecca, Hartley, Wirz and Woll2011). The most widely used of these tests in the literature are the BSL receptive (Herman, Holmes & Woll Reference Herman, Holmes and Woll1999) and productive tests (Herman et al. Reference Herman, Grove, Holmes, Morgan, Sutherland and Woll2004). The receptive skills test has been adapted for the USA context (Enns & Herman Reference Enns and Herman2011), Germany (Haug Reference Haug2012) and Australia (Johnston Reference Johnston2004). Advances in sign language assessment development are also reported in the Netherlands (Hermans, Knoors & Verhoeven Reference Hermans, Knoors and Verhoeven2010). These tests currently only provide fractional information about sign language development. Nonetheless, the outcomes confirm the problems that some deaf children experience in developing sign language fluency (with the exception of deaf children of deaf parents) even for pupils in bilingual educational programmes (Johnston Reference Johnston2004).

2.4 Summary

The study of sign language acquisition provides a foundation for understanding bimodal bilingualism and has changed our understanding of human language learning and the mind. This research has shown that sign languages are linguistically organised and that meaning is produced and perceived in the same way as spoken languages, even though the mode of communication is visual-manual and not auditory-oral. We have learned that the architecture of the brain for the structure and processing of language exists independent of modality and that the milestones for sign language acquisition are parallel to those of spoken language where language is accessible in the learning environment.

This research provides the linguistic context for the exploration of bimodal and bilingual education. From this knowledge base we can appreciate the bilingual language learning potential and challenges for deaf children and the very specific issues for children in hearing families (over 95% of deaf children) who grow up without access to a fluent model of sign language from birth. Application of this knowledge to learning and teaching requires an understanding of actual bimodal bilingual language experience and use to see how the linguistic research might engage with everyday interactions in deaf children's lives.

3.1 Research with hearing bimodal bilinguals

Studies of interaction between hearing children and their deaf parents provide some insights into fluent bilingual and bimodal usage, while acknowledging that different language learning experiences obtain for most deaf children in terms of early language access and fluency. In these case studies deaf parents use both spoken and sign language with their deaf and hearing children (van Den Bogaerde & Baker Reference van Den Bogaerde and Baker2005; Baker & Van Den Bogaerde Reference Baker, Van Den Bogaerde, Plaza-Pust and Morales-Lopez2008). Mixing sign and spoken language is a language contact phenomenon and, as such, a natural part of daily interaction even where sign language is the shared and fluent language for all.

This type of language contact is illustrated further in examples of code blending by adults and children that draws on and integrates the grammars of sign and spoken language. A specific example of this is the presence of ASL in spoken English structures as when the use of grammatically correct ASL facial expression accompanies spoken wh-type questions or the use of pointing during a spoken English utterance clarifies the subject and object of a sentence (Bishop & Hicks Reference Bishop and Hicks2005; Bishop Reference Bishop2010). This way of blending the two languages is again a natural feature of talk in a bimodal bilingual language environment. It is acknowledged here that this is a specific language learning context. Nonetheless, these findings exemplify fluent bimodal bilingual language usage and illustrate that flexible mixing and blending of sign and spoken language constitutes a normal part of individual language repertoire in this context.

3.2 Research with deaf bimodal bilinguals

Research into deaf bimodal and bilingual language usage has only recently begun to emerge and, as with the hearing bimodal bilingual research, these are often small-group or case studies (for example Levesque, Brown & Wigglesworth Reference Levesque, Brown and Wigglesworth2014). The key research questions consider how deaf children use their sign and spoken languages in different situations with deaf and hearing communication partners, how their different and blended language competencies can be measured and described, and how their spoken and sign language repertoires and usage change over time.

The main findings that emerge from this work ascertain that deaf children's alternate and blended use of sign language and spoken language is a normal component of bimodal bilingual communication. Deaf bimodal bilingual children are able to switch flexibly between languages and modes and in some studies the most advanced use of language, in terms of lexical richness and syntactic complexity was when sign and speech were used together (Klatter-Folmer et al. Reference Klatter-Folmer, van Hout, Kolen and Verhoeven2006; Krausneker Reference Krausneker, Plaza-Pust and Morales-Lopez2008; Rinaldi & Caselli Reference Rinaldi and Caselli2009).

In terms of description and assessment, although we have some sign language assessments and many spoken language assessments developed for deaf children, we currently lack tools or protocols to document deaf children's mixed or blended use of sign and spoken language. We also lack systematic means of accurately identifying language dominance. Studies which have sought to explore this have tended to develop context-specific protocols. Examples can be found in research in Brazil and the Netherlands (Klatter-Folmer et al. Reference Klatter-Folmer, van Hout, Kolen and Verhoeven2006; Lichtig et al. Reference Lichtig, Couto, Mecca, Hartley, Wirz and Woll2011). These studies reveal the extent to which deaf children's mixed (signed and spoken) utterances can provide the most efficient means of producing linguistic complexity and communicative richness. Although there is some question as to whether these mixed utterances inhibit children's optimal use of either language, the findings suggest that the separate assessment of languages does not demonstrate the linguistic complexity of which deaf children are capable.

3.3 Research with children with cochlear implants

Deaf children's changing bimodal and bilingual language practices have come under scrutiny in the specific context of CI research because of the early and increased access to the full range of speech sounds that CIs afford. The use of sign language changes following implantation but nonetheless continues, particularly for those children with additional difficulties or where the outcomes of the CI are not as predicted (Hyde & Punch Reference Hyde and Punch2011). Studies of children's language show a general shift towards spoken language use but with the continued flexible use of sign language depending on individuals and contexts (Watson, Archbold & Nikolopoulos Reference Watson, Archbold and Nikolopoulos2006). Parents and young deaf people themselves report that sign language continues to be useful after implantation, providing support for learning, communication and networking flexibility, and for social-emotional development and identity (Christiansen & Leigh 2004; Preisler, Tvingstedt & Ahlström Reference Preisler, Tvingstedt and Ahlström2005; Watson et al. Reference Watson, Hardie, Archbold and Wheeler2008; Wheeler et al. Reference Wheeler, Archbold, Hardie and Watson2009; Walker & Tomblin Reference Walker, Tomblin, Marschark, Tang and Knoors2014). Ongoing research shows that bimodal and bilingual children with CIs do not suffer any disadvantage in the development of spoken language. In fact, early sign language input mitigates the disadvantages of early auditory deprivation (Davidson, Lillo-Martin & Chen Pichler Reference Davidson, Lillo-Martin and Chen Pichler2014). Furthermore, implanted native signers develop better spoken language skills when compared to children with hearing parents (Hassanzadeh Reference Hassanzadeh2012).

An important conclusion drawn from this work is that sign language continues to be part of the communication journey of young people with implants and an important part of growing up deaf with an implant (Christiansen & Leigh 2004; Cramér-Wolrath Reference Cramér-Wolrath2013). This area of research marks a significant development in deaf education's understanding of children's language usage and provides insights into the flexibility and pragmatism of young deaf people and their families. This avenue of research also incorporates the voices of young deaf people and their parents about their language use, rarely visible in the language research to date.

3.4 Summary

The research into bimodal bilingual language experience and use illustrates that the mixed and blended use of sign and spoken language is a normal part of deaf children's daily lives. However, running through this research are questions regarding the extent to which mixed and blended language use compromises or supports the autonomous development of the individual languages. There is an evident tension here in understanding mixed language use: either it is a stage in developing competence in both languages or it is a creative use of new language structures and as such a stable end point in itself. These questions probe theoretical issues regarding the nature of language itself, as separate bounded systems or a dynamic set of resources for meaning-making. The limited evidence provided by case studies observing deaf children thus far suggests an integrated language potential that exceeds the skills in either languages when examined in isolation. These findings encourage a move away from a fractional view of bilingualism which focuses on separate proficiencies in separate languages and from pedagogies which aim to separate the use of sign and spoken language. That is not to suggest that proficiency in both languages does not remain a goal, but rather that we investigate the creative use of children's full language repertoires to achieve such a goal. The development of this new knowledge also highlights the need for more integrated approaches to assessment which recognise the fluctuating dominance of different languages in children's lives and different language proficiencies across different domains.

4.1 The theoretical debate

As sign language became more widely used in education, and research began to develop, a number of studies in the 1980s and 1990s demonstrated a positive relationship between reading skills and sign language competency. This body of research revealed that deaf children of deaf parents, where sign language is the shared language of the home, are generally more successful readers than deaf children of hearing parents (Prinz & Strong Reference Prinz and Strong1998; Hoffmeister Reference Hoffmeister, Chamberlain, Morford and Mayberry2000; Padden & Ramsey Reference Padden, Ramsey, Chamberlain, Morford and Mayberry2000). The finding that good signers are good readers was interpreted as evidence of how sign language could support deaf children's literacy development and the potential of routes to literacy which need not necessarily rely on access to the phonological aspects of the spoken language. A model of bilingual education thus began to emerge in the 1980s which envisaged sign language as the L1 or dominant language and a basis from which to develop literacy skills in the second language (L2). As this model became increasingly articulated, especially in the USA, Scandinavia and the UK, Cummins's linguistic interdependence theory (1981) became an accepted theoretical basis upon which to speculate about the role of sign language in supporting literacy development (Mayer Reference Mayer2009; Grosjean Reference Grosjean2010; Swanwick Reference Swanwick2010).

The ensuing debate problematised the bilingual model of literacy development which assumes sign language to be the L1 and written language to be the L2. In a landmark and influential publication, Mayer & Wells (Reference Mayer and Wells1996) argued that deaf children are not learning the written form of the spoken language in circumstances commensurate with other L2 learners for a number of reasons. Firstly, sign language has no written form and so, unlike most L2 learners, deaf children have no prior experience of the use of this modality. Secondly, deaf children are usually unable to approach the learning of L2 literacy with established and fluent L1 sign language skills. Deaf children are thus not able to approach the learning of literacy as an L2 or an L1 experience.

Mayer & Wells (Reference Mayer and Wells1996) described this as a ‘double discontinuity’ (p. 103) and their theoretical critique has been influential in the deaf education literature. However, this critique is problematic in a number of ways. Firstly, the argument that transfer cannot take place between dissimilar languages is based on a narrow view of linguistic transfer and of routes to literacy development. Cummins's (2006) view of linguistic transfer incorporates conceptual knowledge, metacognitive strategies and pragmatic skills as well as linguistic and phonological knowledge. There is growing evidence of positive interaction between sign and spoken languages at all of these levels and indeed that transfer can take place across between two or more dissimilar languages (Tang, Lam & Yiu K Reference Tang, Lam, Yiu K, Marschark, Tang and Knoors2014).

Further, this critique assumes that there are no goals relating to spoken language development in bimodal bilingual deaf education and that the only route to literacy considered is via a sign language. This ignores the fact that approaches to bilingual education are more diverse and nuanced than this. The goals associated with bilingual education and interventions to support and enhance spoken language development are not mutually exclusive (Swanwick et al. Reference Swanwick, Dammeyer, Hendar, Kristoffersen, Salter, Simonsen, Marschark, Tang and Knoors2014).

The theoretical debate demonstrates the need for a model of bimodal bilingual literacy development that draws on research in the wider field of bilingualism while remaining relevant to deaf learners. This involves understanding how bimodal bilingual deaf children deploy their language resources in the process of learning to read and write, and examining bimodal bilingual pedagogies that have been developed to support these processes.

4.2 The reading processes

Research into bimodal bilingual deaf children's language mixing and blending has been used in the exploration of deaf children's literacy development. One aspect of this research literature is concerned with the neurological processes involved for bilingual deaf readers, recognising that the reading process may be a qualitatively different experience for these learners. For hearing unimodal bilinguals we know that both languages are activated in reading, whether or not the two languages have similar scripts (Piñar, Dussias & Morford Reference Piñar, Dussias and Morford2011). That is to say that there is evidence of cross language activation between words which are phonologically or orthographically similar. The case of bimodal bilinguals raises the question as to whether cross-language activation happens during reading, given that the two languages do not share phonology or orthography. This research has looked at the extent to which written words activate sign translations in deaf (proficient) bimodal bilinguals. Morford & Hänel-Faulhaber (Reference Morford and Hänel-Faulhaber2011) present evidence that deaf bilinguals activate the sign translation of words when processing written words. This evidence of cross-language processing is also reported by Ormel et al. (Reference Ormel, Hermans, Knoors and Verhoeven2012) who conducted a similar experiment with deaf children (who were not yet proficient bilinguals) using word picture pairs with and without strongly iconic sign translation equivalents. These findings mirror those for less proficient hearing unimodal bilingual children, who activate their L1 when reading.

4.3 Language interaction

The interrelationship between sign, spoken and written languages has also been explored in the domain of bimodal bilingual literacy development. There is emerging empirical evidence aimed at the lexical level, and particularly at vocabulary learning, that found children with more extensive sign language vocabularies tend to have more extensive written vocabularies (Hermans et al. Reference Hermans, Knoors, Ormel and Verhoeven2008; Holzinger & Fellinger Reference Holzinger, Fellinger, Marschark, Tang and Knoors2014). This work has made an important contribution to our thinking about transfer between sign and spoken languages, suggesting that transfer at the semantic lexical and grammatical level can be obtained in the case of bimodal bilingual deaf learners (Tang et al. Reference Tang, Lam, Yiu K, Marschark, Tang and Knoors2014). Hermans, Ormel & Knoors (Reference Hermans, Knoors and Verhoeven2010) suggest that this should be described as an ‘associative’ and not a ‘direct’ transfer, but argue that this can nonetheless can be ‘cultivated’ (p. 194) through the right kind of teaching to become automatic.

The pedagogical question that this research raises is how to support learners to make the association between sign and written vocabularies. Hermans et al. (Reference Hermans, Knoors, Ormel and Verhoeven2008) propose that one way to do this is through ‘chaining’. This term refers to a way of making connections between ASL sign, fingerspelling and written English (Padden & Ramsey Reference Padden, Ramsey, Chamberlain, Morford and Mayberry2000). Typically, it uses different sequences of the presentation of a sign, the fingerspelling of its English equivalent and the written version of the same word. This can be done in any order, depending on the teaching objective, but the strategy is often used where there is a teaching emphasis on learning new English terminology or vocabulary or to point out equivalence of meaning across two languages. For example, a teacher might point to a new English word, fingerspell the word and then give the ASL sign for the word. The term ‘sandwiching’ is also used for this type of language presentation, which implies the alternated use of signs and fingerspelt words. In much the same way as spoken and written languages are blended in reading (we say the words as we read or write them), encoding written words via a sign-based system cultivates links between the two language modalities (Mayberry, DelGiudice & Lieberman 2011).

Hermans et al. (Reference Hermans, Knoors, Ormel and Verhoeven2008) also emphasise that the role of spoken language, specifically phonological awareness, should not be overlooked in relation to bilingual deaf children's reading development. Some of the sign based support systems such as visual phonics and cued speech and speechreading, which exploit associations between signs and spoken and written words, have a role to play here (Wauters et al. Reference Wauters, Knoors, Vervloed and Aarnoutse2001; Mollink, Hermans & Knoors Reference Mollink, Hermans and Knoors2008; Harris & Terlektsi Reference Harris and Terlektsi2011). These are discussed in full in Section 3.6.

4.4 Bimodal bilingual deaf children's writing

The study of deaf children's writing has provided fertile ground for analysing children's bimodal bilingual usage and for speculation about the interaction between visual/gestural and written modalities. The challenges that this experience of language learning presents for writing centre on the ‘unmappability’ between multi-channel visual languages and written languages, such as the absence of one-to-one correspondence between written and spoken words and signs and the different grammatical structures of the languages (Swanwick, Kitchen & Clarke Reference Swanwick, Kitchen and Clarke2012). The most fruitful studies in this area are those which consider deaf children's writing from a bilingual rather than monolingual perspective and explore the different potential routes to literacy development, documenting how children use their repertoire of bilingual and bimodal resources (Swanwick Reference Swanwick2002; Singleton et al. Reference Singleton, Morgan, DiGello, Wiles and Rivers2004; Koutsoubou, Herman & Woll Reference Koutsoubou, Herman and Woll2006, Reference Koutsoubou, Herman and Woll2007; Niederberger Reference Niederberger, Plaza-Pust and Morales-Lopez2008; van Beijsterveldt & van Hell Reference van Beijsterveldt and van Hell2012).

From this perspective, the concept of transfer between languages has been explored via the influence of sign language on writing and the extent to which children draw on the resources of the visual-spatial language that they know to support the writing process across differences in modality. Menéndez (Reference Menéndez2010) describes deaf children's use of sign language structures in their writing as a transfer strategy. He suggests that that they draw on the skills that they have in sign language to engage with a written language and that the route to literacy does not have to be a predominantly phonological pathway. Niederberger (Reference Niederberger, Plaza-Pust and Morales-Lopez2008) argues that linguistic transfer can be identified at a pragmatic level, demonstrating that knowledge of construction and comprehension of narratives in sign language is transferable to deaf students’ performance in their writing. Van Beijsterveldt & van Hell (Reference van Beijsterveldt and van Hell2010, Reference van Beijsterveldt and van Hell2012) suggest that the enhanced use of evaluative writing devices by good sign language users is also evidence of language transfer, which reflects how integral the communication of evaluative information about objects or people (emotion states) is to description in sign language. Swanwick (Reference Swanwick2002) proposes that transfer can be construed at a metalinguistic level, which showed that deaf children's awareness of the different ways in which sign language and written languages express meaning supported the writing process. Koutsoubou's (2007) finding, that writing from a translation task facilitated more complex writing and narrative structure, supports this thesis.

4.5 Bimodal bilingual approaches to supporting literacy development

Research into bimodal bilingualism and literacy development adds to our understanding of linguistic interconnections between languages and the flexibility of the language system in accommodating and working between different modalities. However, there is as yet an insufficient amount of research on which to build evidence-based literacy pedagogies for deaf bimodal bilingual learners (Myers et al. Reference Myers, Clark, Musyoka, Anderson, Gilbert, Agyen and Hauser2010; Cannon & Guardino Reference Cannon and Guardino2012). The literature in this area focuses more on development than on the evaluation of teaching approaches addressing the ‘double discontinuity’ described by Mayer & Wells (Reference Mayer and Wells1996). Opinion is divided regarding the extent to which bilingual deaf readers do, or do not, rely on auditory phonological awareness (e.g. Chamberlain & Mayberry Reference Chamberlain, Mayberry, Chamberlain, Morford and Mayberry2000; Wang et al. Reference Wang, Trezek, Luckner and Paul2008; Miller, Lederberg & Easterbrooks Reference Miller, Lederberg and Easterbrooks2013). Furthermore, the correlation between speechreading and subsequent literacy skills suggest that phonological awareness does not have to be a wholly auditory process (Kyle & Harris Reference Kyle and Harris2011). Nonetheless, several avenues of research have explored ways to ‘bridge the gap’ between sign language and literacy by using multi-sensory coding to support word learning and reading comprehension (van Staden Reference van Staden2013).

These lines of enquiry are underpinned by the predominant perspective that literacy development relies on some conversational experience of the target language and an internal representation of the spoken language (Hermans et al. Reference Hermans, Knoors, Ormel and Verhoeven2008). This raises the pedagogical question: how may the rich oral and conversational experience needed to support the development of text based literacy skills be facilitated? This question has prompted the exploration of teaching approaches which deploy the mixed and blended use of sign language and spoken language.

4.6 Mixed and blended language use in literacy development

Most of the terms related to mixed and blended use of signs and speech used in this review emanate from the USA or the UK literature on the various invented sign systems which have been developed to represent English semantics and syntax and which have a specific pedagogical purpose relating to the teaching of English.

Manually Coded English (MCE) is a general term which encompasses all the different ways in which sign language and spoken language can be used in combination to provide visual support for spoken English. In broad terms, MCE involves the use of signs borrowed from BSL or ASL (usually meaning-carrying words) to accompany speech. This should not be confused with Total Communication (TC), a term which originally evolved to describe an educational philosophy and approach involving the flexible use of sign and spoken language to meet individual communication needs (Evans Reference Evans1982). Over the years the philosophical use of this term to denote a child-centred approach has been lost and TC is now more often used to describe the simultaneous combination of signs with speech (Marschark, Knoors & Tang Reference Marschark, Knoors, Tang, Marschark, Tang and Knoors2014). This is generally referred to as Simultaneous Communication (SimCom) in the American literature, and Sign Supported English (SSE) in the UK. All of these terms refer in some way to the simultaneous production of grammatically correct spoken English and signs for educational purposes and/or for a mixed deaf/hearing audience.

4.7 Insights from early literacy

Alongside these artificial mixed and blended approaches to supporting literacy development, research into how deaf children read and share books with deaf parents offers insights into natural and fluent bimodal bilingual interaction around written language. This work began to emerge in the 1980s (Swanwick & Watson Reference Swanwick and Watson2005) and there have been several studies since 2000 which have looked specifically at ways in which bilingual deaf parents (usually mothers) mediate between written language and sign language during book-sharing activities with their young deaf children (Plessow-Wolfson & Epstein Reference Plessow-Wolfson and Epstein2005; Kaderavek & Pakulski Reference Kaderavek and Pakulski2007; Swanwick & Watson Reference Swanwick and Watson2007; Mueller & Hurtig Reference Mueller and Hurtig2010; Berke Reference Berke2013).

Some of the techniques that mothers reportedly use reflect those described in the pedagogical literature, such as the use of sign language to talk about language and specifically to give definitions, and interpret and explain either individual words or written language conventions and features such as rhyming or onomatopoeia. Parents also use chaining to pair words to signs through fingerspelling, and there is evidence to suggest that very young deaf children initially see fingerspelt words as a single integral shape. By the age of five, they start to see that there is an individual correspondence between a fingerspelt and a written letter (Padden Reference Padden, Schick, Marschark and Spencer2006). Another strategy is parents’ use of signing in English word order or incorporating English grammatical features in a signed utterance (such as a fingerspelt function word), which is also documented as a teaching strategy in the pedagogical research literature (Humphries & MacDougall Reference Humphries and MacDougall2000). Aside from these linguistic practices, these studies observe the setting, positioning and placement of the signing by deaf parents vis-à-vis the child and the book, and how they manage this along with touch and non-manual communication to engage and maintain attention and indicate shifts in the activity, add characterisation or emphasis. This research has to some extent brought the voice of deaf parents of deaf children to the fore. The value of the parent experience and perspective has been recognised as essential to understanding bilingual literacy development and to underpin coaching and support strategies for hearing parents and teachers (Mounty, Pucci & Harmon Reference Mounty, Pucci and Harmon2014). However, there are as yet few published studies which provide evidence of positive outcomes of coaching other than increased reading time and effectiveness of communication (Delk & Weidekamp Reference Delk and Weidekamp2001). Nonetheless, we have much to learn from this research about how sign and spoken language might be used to support literacy development and how parents facilitate this.

4.8 Summary

The research questions related to bimodal bilingualism and literacy span theoretical and pedagogical issues about the interaction between sign, spoken and written modalities. The theoretical questions have centred on the interrelationship between signed, spoken and written communication. This work has contributed to, and expanded, our understanding of routes to literacy development and in particular the different ways in which phonological knowledge can be acquired and processed. The outcomes of this research have also extended our understanding of language transfer by demonstrating that, regardless of difference in modality, transfer can occur at the conceptual, metalinguistic, linguistic and phonological levels as Cummins (Reference Cummins2006) suggests in his treatment of unimodal bilingualism.

In terms of pedagogy, this area of research has provoked debate about approaches to bimodal bilingual education for deaf children. The role of spoken language and the various mixed and blended language systems have been extensively discussed and described. However, empirical evidence demonstrating the success of any one approach over another cannot be found. This is hardly surprising, given the heterogeneity within any group of deaf learners of language experience, use and ability. What emerges from this research is that understanding the individual language profiles and skills within and across sign and spoken language must precede decisions about language policy. It is clear that teachers need an extensive repertoire of specialist teaching skills and strategies for making links between sign and spoken/written languages in order to support bimodal bilingual deaf children's literacy development effectively (Wauters & de Klerk Reference Wauters, Klerk, Marschark, Tang and Knoors2014). Moreover, the evidence base for these practices needs to be developed to provide a steer for education and training and ensure that such skills are properly deployed.

5.1 Language use in the classroom

Discussions of practitioner use of sign language and spoken language in the teaching context usually centre on text-based activities where the role of sign language is explored as a bridge to facilitate the learner's access to, and understanding and use of, written language (Akamatsu, Stewart & Becker Reference Akamatsu, Stewart and Becker2000). Much of this work is concerned with SimCom in bilingual education, trying either to critique or defend its use, and the writing here tends to be passionate (Tevenal & Villanueva Reference Tevenal and Villanueva2009). This is explained by a legacy of anxiety about the use of SimCom in deaf education practice, which has both linguistic and ideological roots, as discussed in Section 3. What comes through from this body of literature is that language in the classroom has thus far been driven by individual beliefs, attitudes and the language skills of educators in this field, rather than being responsive to children's usage.

Despite these language politics, there are studies which have tried to evaluate the specific role of English-based signing in improving deaf children's English learning and demonstrate the impact of bimodal bilingual pedagogical approaches. This body of work provides some description of the creative ways that teachers blend or mix sign and spoken languages to provide lexical, semantic and conceptual support for learners who may not be fluent signers or speakers. Some of these emphasise the importance of separating the languages in teaching. In these cases ASL is described as the main language of instruction, and access to English is facilitated via literacy with little focus on spoken language development. Although the teachers switch between languages, the blended use of voice with signs (SimCom) is discouraged (DeLana, Gentry & Andrews Reference DeLana, Gentry and Andrews2007; Andrews & Rusher Reference Andrews and Rusher2010). Other studies have specifically explored the effectiveness of English-based signing in providing an internal representation of English in order to support literacy development (Mayer & Akamatsu Reference Mayer and Akamatsu2000; Power, Hyde & Leigh Reference Power, Hyde and Leigh2008). The detail given about the way in which language is actually used in the classroom is very variable and in some cases the learning context is described very generally as a Signed English or dual language environment. In other examples, bimodal bilingual classroom strategies are documented in detail. Among these, Humphries & MacDougall (Reference Humphries and MacDougall2000) provide an informative taxonomy of strategies used by deaf and hearing teachers demonstrating the practice of ‘chaining’ as one of the ways of making connections between ASL signs, fingerspelling and written English.

Methodologically, all of these studies struggle with the difficulties of demonstrating the effectiveness of one or other approach. The Andrews & Rusher (Reference Andrews and Rusher2010) study is informative in terms of teacher practices, but, from the data provided, we cannot conclude that it is code switching specifically that makes a difference to children's learning. The Delana et al. study (Reference DeLana, Gentry and Andrews2007) shows a relationship between number of years of ASL use and reading comprehension and raises interesting points about early language use and parental involvement, but there is no clear link between the instructional practices and the learning outcomes. Similarly, no direct link can be made in the Mayer & Akamatsu study (Reference Mayer and Akamatsu2000) between the use of English-based signing, grammar skills and reading improvement other than to say that the use of English-based signing improves deaf children's English-based signing and does not ‘hurt’ literary development (p. 462).

Some bilingual teaching approaches are still based on the perceived need to separate languages in the teaching environment to preserve the integrity of each language. Andrews & Rusher's (Reference Andrews and Rusher2010) study, for example, presents a very ‘monolingual’ view of deaf children's language potential with an emphasis on switching between the two languages to give learners the language that they need to ‘survive’ (p.420). The Delana et al. study (Reference DeLana, Gentry and Andrews2007) also emphasises a ‘dual language methodology’ (p.74), where the use of SimCom is avoided. Research is developing, however, to examine more closely the role of English-based signing such as SE and SSE in the teaching and learning process (Knoors & Marschark Reference Knoors and Marschark2012).

5.2 Evidence of learning

Although the emphasis in the pedagogical literature is on mode of delivery, interest is developing the quality of interaction around learning and the actual learning taking place. Molander, Halldén & Lindahl (Reference Molander, Halldén and Lindahl2010), for example, look at the difference between deaf and hearing learners’ extended discussions around a learning objective and in particular their scientific reasoning. They found that deaf learners’ discussions are more quickly truncated than those of their hearing peers because of difficulties with new curriculum vocabulary and nuanced meanings of certain curriculum terms. Showing the importance of looking beyond modality, they demonstrate that, although new vocabulary can be presented in fingerspelt form, this does not necessarily connect to meaning generally for deaf learners or the appropriate meaning for a given learning context. These difficulties were found to inhibit deaf and hearing group work and thus curtail deaf learners’ opportunities to extend learning through discussion with peers.

The Kelman & Branco (Reference Kelman and Branco2009) study of deaf and hearing teaching teams in an inclusive setting also examines interaction in the classroom. They look beyond modality to consider what makes for successful dialogue in the learning context and how the use of non-verbal features such as touch, proximity, and use of space in the classroom supports mutuality, co-construction and negotiations of meanings.

Marschark et al. (Reference Marschark, Leigh, Sapere, Burnham, Convertino, Stinson, Knoors, Vervloed and Noble2006) look beneath the surface issues of delivery to examine access to and comprehension of bimodal and bilingual instruction. They compared the accessibility of information provided to students in ASL and MCE, and explored the effects on comprehension when real time text is added into the delivery. Their study of young deaf adults (post 16) provides evidence that adding sign language support to lesson delivery does not ensure deaf students’ learning or comprehension to a level equivalent to that of their hearing classmates (Marschark et al. Reference Marschark, Leigh, Sapere, Burnham, Convertino, Stinson, Knoors, Vervloed and Noble2006; Marschark et al. Reference Marschark, Sapere, Convertino, Mayer, Wauters and Sarchet2009). Although some combinations of information sources, such as the addition of text, certainly enhance access, deaf students still learn less than hearing peers.

This work illustrates that bimodal bilingual delivery in the classroom, even from skilled sign language users, does not in itself guarantee learning. More research is needed into the quality of interaction in bimodal bilingual teaching and how deaf and hearing teachers can combine bimodal bilingual teaching tools with strategies that extend discussion and facilitate engagement and learning.

Work which has examined teaching ‘quality’ and how it might be identified and measured in a bimodal bilingual context reiterates this finding (Knoors & Renting Reference Knoors and Renting2000; Hermans et al. Reference Hermans, Wauters, De Klerk, Knoors, Marschark, Tang and Knoors2014). These studies identify the language of instruction, the interactional style of the teaching, student engagement and teacher-student relationships as indicative of the quality of instruction. Using these measures, and incorporating the view of the learners, Hermans et al. (Reference Hermans, Wauters, De Klerk, Knoors, Marschark, Tang and Knoors2014) suggest that teachers still find it difficult to attune the language input and classroom talk to bimodal bilingual deaf learners and engage them fully in learning. This focus on learning and teaching illustrates that bilingual programmes may need to adapt themselves to the changing communication profiles and learning needs of deaf children. In response to a changing language landscape, bimodal bilingual approaches need to evolve and embrace the diverse language learning profiles of deaf children. Specifically, there is a need to closely examine the way in which sign and spoken/written languages are used in the classroom and how different types of language use support learning. Developing bimodal bilingual pedagogies entails a review of what we know and what we can learn from other bilingual contexts, as well as revisiting some of the essential conditions that facilitate learning in the classroom.

5.3 Outcomes of bilingual education

There are repeated calls in the literature for the success of bilingual education to be demonstrated in terms of academic outcomes. The central criticism is that no research demonstrates that it is a more successful approach than any other, in terms of academic attainment. This critique has fuelled a fruitless and problematic dialogue on a number of counts. The first is that bilingual education was not established with the goal of being a more successful approach than any other. Instead, it developed as an educational response to the language and learning needs of deaf children, for whom spoken language approaches are not sufficiently inclusive. Secondly, the measurement of the success of an educational approach in terms of academic outcomes has eclipsed the discussion of the wider and fundamental linguistic, cognitive and social-emotional outcomes. Thirdly, to produce research with a small and heterogeneous population that demonstrates the superiority of one educational approach over another is fraught with methodological issues. The ‘weak’ support in the literature is not limited to the bilingual approach. There is no educational approach in deaf education that has been identified as more successful than any other.

Studies of the academic outcomes of bilingual education are sparse and generally inconclusive, even in settings where bilingual education is well established. The available national studies tend to demonstrate limited educational advantages and do not suggest that this approach closes the progress gap between deaf and hearing children (Hendar Reference Hendar2009; Rydberg, Gellerstedt & Danermark Reference Rydberg, Gellerstedt and Danermark2009).The methodological issues centre on the difficulty of demonstrating a direct relationship between a bilingual approach and children's learning outcomes. Although some relationships are found between children's attainment and background factors such as language experience and use, parental involvement, level of hearing loss, IQ, ethnicity and use of amplification, they are not causal and do not thus link attainment to approach. Further, many of the studies available are largely descriptive of bilingual policy and practice and lack established measures of interventions or teaching approaches, and pupil data often lacks sign language or spoken language measures. It is also to be noted that few of these studies have appeared in peer-reviewed journals.

There are a few studies that show the learning achievements of students in specific bilingual contexts and, as such, provide useful programme evaluation information and some evidence of attainment against national benchmarks. Lange et al. (Reference Lange, Lane-Outlaw, Lange and Sherwood2013) and Nover et al. (Reference Nover, Andrews, Baker, Everhart and Bradford2002) are both examples of longitudinal studies of cohorts of pupils in bilingual schools for the deaf in the USA. Lange et al. (Reference Lange, Lane-Outlaw, Lange and Sherwood2013) demonstrate that after four years ‘immersion’ in a bilingual programme, 66% achieve average or above average scores in maths and reading, comparable to the nationally normed group of deaf and hearing children. Nover et al. (Reference Nover, Andrews, Baker, Everhart and Bradford2002) also found reading comprehension levels above the reported norms for deaf children after at least three years in the programme. We are not able to extrapolate much about the teaching approach from either of these studies in order to analyse these results further and we cannot generalise the findings to other deaf learners. They nonetheless present evidence of the benefits of this approach for some deaf children.

Aside from these isolated studies of attainment, there are some reported outcomes of bilingual education with regard to cognitive control and metalinguistic abilities. This generally relates to the interaction between sign and spoken language and the potential convergence or ‘pooling’ of language knowledge and abilities (Mineiro et al. Reference Mineiro, Nunes, Moita, Silva, Castro-Caldas, Marschark, Tang and Knoors2014; Ormel & Giezen Reference Ormel, Giezen, Marschark, Tang and Knoors2014; Plaza-Pust Reference Plaza-Pust, Marschark, Tang and Knoors2014). Other identified advantages include the psychosocial wellbeing resulting, in part, from improved access to the curriculum and to peer interaction in the classroom (Bagga-Gupta Reference Bagga-Gupta2000; Dammeyer Reference Dammeyer2010). The establishment of early language and communication (Rinaldi et al. Reference Rinaldi, Caselli, Onofrio, Volterra, Marschark, Tang and Knoors2014) as well as support for the development of reading and vocabulary knowledge (Hermans et al. Reference Hermans, Knoors, Ormel and Verhoeven2008) and early literacy skills (Kristoffersen & Simonsen Reference Kristoffersen and Simonsen2012; Dammeyer Reference Dammeyer2014). These more general gains can be explained by the advantages that fluent, early communication provides in terms of linguistic, cognitive and social-emotional development, but they have yet to translate into attainment data.

5.4 Summary

This review illustrates the range of bimodal and bilingual instructional practices deployed by teachers in deaf education. They signify a move towards examining the outcomes of bimodal bilingual teaching approaches but they are limited in terms of their scope and methodological rigour. What emerges from the review of practice is a predominant focus on language modality issues, with the focus on policy rather than practice. There is a gap in the research relating to actual learning in bilingual settings and the detail of teaching approaches. We need to know more about how L1 and L2 teaching methods are differentiated and employed and how languages are used in the classroom and across the curriculum.

Furthermore, the intricacies of communication around learning in the classroom are seldom explored in terms of learning and teaching styles, interaction, dialogue and engagement in the learning context, and how that is challenged and optimised by bilingual and bimodal methods. There are very few studies which look at curriculum specific areas beyond literacy. There is emerging treatment of science, technology and maths subjects but not specifically with reference to bilingual education, and there is developing interest in the use of technology in bilingual contexts. In terms of overall outcomes, the results are mixed. The use of sign language clearly supports some aspects of learning and development for some children but has not closed the progress gap between deaf and hearing children. The confounding factors in the research make it difficult to extrapolate the nature of the support for learning that sign language provides and the relationship between experience, environment and educational outcomes.