Shorter but richer versus longer with less information: linguistic differentiation between British Sign Language and sign supported English

2.1 British Sign Language

BSL is a natural language used by Britain’s deaf communities. To fully understand different varieties of BSL, including the extent of English influence on BSL, it is necessary to delve into deaf education. As 90–95 % of deaf children are born to hearing families (Mitchell and Karchmer 2004), deaf children are likely to get most of their sign language input in educational settings. This means that deaf children’s educational placements have (and have always had) a huge impact on their sign language development.

The emergence of BSL as we know it today is often associated with the establishment of deaf schools in the 18th and 19th centuries (Brown 2020; Quinn 2010; Sutton-Spence and Woll 1999). This is how it is thought that most sign languages emerge, when deaf people form a community as in the context of the beginning of deaf education (cf. the recent case of Nicaraguan Sign Language; Kegl et al. 1999). The first school for deaf children in the UK was set up in Edinburgh, Scotland by Thomas Braidwood in 1760. There is some evidence that ‘deaf signing’ (as it was known at the time) was used as a method of instruction; however, it is likely that the structure of signing followed the order of English (Kyle and Woll 1985). In 1792, Joseph Watson, the nephew of Thomas Braidwood (who worked at Braidwood school), set up the first public school for deaf children in London. Joseph Watson pushed for ‘natural signing’ (Watson 1809), which seems to refer to deaf children’s natural sign language communication without pushing them to incorporate features from English, including structure.

For about over 100 years, from 1760, sign language was used for educating deaf children (likely to varying degrees) in the UK. By 1870 there were 22 schools for deaf children around the UK, many of them residential and many with former deaf pupils who later became teachers (Kyle and Woll 1985). At the International Congress on the Education of the Deaf held in Milan 1880, sign language was ‘banned’ in deaf education, in favour of educating deaf children using spoken language (known in deaf communities as ‘oralism’).

From the Milan conference in 1880 to the 1970s, deaf children were rarely taught, if ever, in sign language. What Joseph Watson called ‘natural signing’ (likely the precursor to modern BSL) was likely to be largely restricted to social gatherings outside of educational settings, e.g., at deaf clubs or in deaf, signing families. There were fewer deaf teachers during this period and less use of signing in the classroom (Brennan 1992). English and English-based signing were given more prestige. ‘Deaf signing’ was considered inferior, even by deaf BSL signers themselves, who formed internalised, negative attitudes about sign languages, after a century of linguistic oppression in education (Ladd 2003).

In the late 1970s, there were two main developments that affected the use of signing in deaf education. The Warnock Report (1978) recommended that deaf children be sent to integrated schools alongside hearing children. This resulted in the closure of many deaf schools in favour of mainstream education. Also, the Conrad Report (1979) found that deaf 16-year-olds were leaving school with a reading age equivalent to 9-year-old hearing children. This pushed schools with deaf pupils across the UK to encourage the use of speech with (some) BSL signs simultaneously (known as “Total Communication”), believing this would enable better English and improve reading (Ladd 2003). However, hearing educators still held almost no respect for BSL, as they believed signed languages are “lesser-than” therefore not deserving of the same respect given to spoken languages. Negative language attitudes played a large role in the decimation of direct education in signed languages (O’Brien 2021).

Towards the late 1970s and early 1980s, BSL started to gain more prestige for various reasons, starting with the recognition of BSL by linguists as a natural language (Brennan 1975). Bilingual education (BSL/English) began in the 1980s which further increased the status of BSL (Swanwick and Gregory 2007). In 2003, BSL was ‘recognised’ by the British Government as the fourth indigenous language used in Britain.^[1] Even with this recognition (and subsequently the BSL Act Scotland in 2015 and UK BSL Act in 2022), BSL is still not seen as having the same level of prestige as English. Its use in education was discouraged until the 1980s. Even today, families with deaf children are still being advised not to sign with their children (Humphries et al. 2017; Rowley et al. 2022).

The systematic historical and ongoing oppression of BSL has had an impact on how it is used. For example, there are clear differences between BSL signers of different ages in relation to fingerspelling, mouthing and lexical signs, which reflect changes in educational policy (Brown and Cormier 2017; Proctor and Cormier 2023; Stamp et al. 2014).

Although BSL is a different language from English, it remains a minoritised language in constant contact with English. As a result, signers often incorporate English features such as mouthings^[2] and fingerspelling (a manual representation of the English alphabet). Also some elements of English and BSL can be produced simultaneously, known as language mixing which is unique to sign languages, e.g., silent mouthing of English words while producing signs; a process known as ‘code-blending’ (Emmorey et al. 2008). However, the extent to which BSL signers use linguistic elements from English varies substantially, dependent on numerous factors including the signer’s speaking/writing ability in English as well as the perceived abilities of their interlocutors (Sutton-Spence and Woll 1999). This high degree of variability is one reason why we have more recent research arguing for translanguaging i.e., use of switching between and mixing of various semiotic resources from different languages and/or modalities – rather than just code-blending between a signed and spoken language (De Meulder et al. 2019; Kusters 2019; Proctor and Cormier 2023). Henner and Robinson (2023) pointed out that translanguaging reflects the reality of how people use languages in opposition to fixed ideas about how languages should function.

BSL is influenced by the majority language (English) for various reasons including oppressive education policies, as described above, which also relates to the power imbalance between BSL and English (Palfreyman and Schembri 2021). Importantly, BSL is not simply English on the hands. Many linguistic descriptions of BSL have demonstrated how its structure differs from English. For example, the order of signs is often different to English, with some structures typologically unique to sign languages. Examples include depicting signs^[3] and indicating verbs,^[4] and taking advantage of simultaneity and spatial modification for grammatical purposes (Cormier et al. 2012; Schembri et al. 2018). There are also mouth gestures i.e., mouth patterns that are not related to spoken language (Boyes Braem and Sutton-Spence 2001).

2.2 Sign Supported English (SSE)

When signers use the term ‘SSE’, they are referring to a contact variety of signing that involves a mixture of BSL and English, with English as the matrix (dominant) language. This usually means that signed utterances follow English grammar but use BSL signs. However, distinguishing between BSL and SSE can be difficult because of constant language contact between BSL and English. Sutton-Spence and Woll (1999) note that with SSE, the “key words of a sentence are signed, while the person speaks. This means that the main vocabulary is produced from BSL, but much of the grammar is English on the mouth” (p. 16). If English is very dominant, the resulting production may more closely resemble spoken English with co-speech gestures resembling or identical to BSL signs.

It is difficult to pinpoint the origins of the term “Sign Supported English”. However, it is quite likely that it is a variety that has come about from extensive language contact reflecting translanguaging practices used by deaf signers. Historically, many children at deaf schools were born hearing and became deaf later (Brown 2020), suggesting many already had some proficiency in English; however, with their deaf peers, they would sign, suggesting code-blending/translanguaging. In the 1970s, when linguistic research on BSL began, linguists believed that spoken and sign language existed on a diglossic continuum and that SSE was somewhere in the middle of that continuum (Deuchar 1977).

A separate but related concept to SSE is Signed English: English-based signing i.e., speaking in English with signs (including signs for English functors) used simultaneously. This method was used in the U.S. using ASL signs (Signing Exact English i.e., SEE – cf. Supalla and McKee 2002). Specific signs were invented for certain words in English not available in ASL e.g., ‘was’ and verb endings (‘ing’, ed’ etc.). There is no equivalent to SEE (i.e., formal creation of signs for function morphemes) in the UK. This may be partly due to early criticism of English-based signing by Joseph Watson, the headmaster of England’s first non-fee-paying deaf school, because of his pushing for “natural signing” (Watson 1809), as noted above. Sutton-Spence and Woll (1999) argued that Signed English differs from SSE in that it incorporates signs that represent grammatical markers such as articles and past-tense markers. So, although signs for English grammatical functors were not explicitly created by educators like they were in the US for SEE, it seems that they may have emerged naturally, particularly in education settings in the UK.

There are several problems with systems relying on signs for grammatical markers like Signed English does. They can become longer and therefore unnaturally slow, with many letting speech take over while dropping some signs (Johnston and Schembri 2007; Ladd 2003; Sutton-Spence and Woll 1999; Supalla and McKee 2002). Another issue is that it is often not possible to match English and BSL lexically. Many phrases or sentences in English can be represented by a single sign due to BSL’s rich morphological system e.g. ‘I ask you’, ‘drink from a mug’ etc. (Johnston and Schembri 2007; Sutton-Spence and Woll 1999). Additionally, some English words are represented by 2 or more BSL signs such as ‘weapon’, ‘multilingual’ etc. Sutton-Spence and Woll (1999) state that traditionally, Signed English is not used in regular communication but only in classes for children to learn English – and that it may be beneficial for learners because it is easier for their hearing teachers.

6.1 Representation in the BSL corpus

Before proceeding, we want to highlight some issues on participant representation in the BSL Corpus. We aimed to recruit deaf people who learnt to sign in their first few years. However, many deaf people learn to sign later, and thus the BSL Corpus does not reflect the full picture of BSL signers. In relation to ethnicity, during our 2008–2010 recruitment drive, we aimed to ensure 10 % of the sample were from ethnic minorities (around 8 % in our corpus). The 10 % aim was based on the census of England and Wales (Office for National Statistics 2011), showing 10 % of the general population were from ethnic minorities (mostly Afro-Caribbean and South Asian). However, this is now considered too low, as it was 20 % in the 2011 census (Wilson et al. 2015). We highlight these issues to recognise the importance of balance and representation in sign language datasets.

6.2 Data annotation

Data annotation consisted of first translating the data from BSL into written English using an English translation tier in ELAN (Wittenburg et al. 2006). Translations were done by qualified BSL/English interpreters and translators and checked by first author Rowley. We then annotated prominent themes in the interview translations, again using ELAN. Interview responses from each participant were coded in a single ELAN file, but with a view of their interlocutor for context (see Figure 2 for example). Responses from the interlocutor were translated and coded in a separate ELAN file. After exporting themes to Excel, we conducted a thematic data analysis, following Braun and Clarke (2006). This involved six stages: (1) data familiarisation, (2) generating some initial notes and codes for emergent themes, (3) searching for those themes, (4) collating those themes together, (5) searching and reviewing themes and then (6) reporting analysis of those themes. Data was analysed by the first author. Given her deaf background, this helped us to understand in detail some of the responses e.g., their educational backgrounds.

Figure 2:

Example of annotations within a single ELAN file. The FreeTransl tier refers to English translations of the participants’ BSL responses.

In the first stage of thematic analysis, general notes were generated highlighting patterns that emerged from the data. For the next two stages, each interview response, and general notes from stage one was examined to develop themes that emerged from the data (stages two and three). Themes were not predetermined and were identified as they emerged from the participants’ responses. Once themes were assigned, the data was exported to Excel and themes were reviewed and re-analysed to ensure each theme was clearly demarcated (stages four and five). Finally, a summary outlining each theme was written up (stage six). Exporting of the data from ELAN to Excel after coding and reviewing themes enabled the grouping of statements according to each theme.

7.1 Structure

Structure was the most prevalent theme and made up 58 % of the tagged data. Structure is discussed according to several sub-categories (e.g., sign order, omission, length and speed, visual expression, mouthing and fingerspelling).

7.2 Context of use

The next theme, context of use, was frequently discussed. Participants gave examples of when SSE would be used and who they believed used SSE. The main domains associated with SSE users were oral/mainstream education and hearing people. These domains often overlapped with one another. In contrast, the use of BSL was associated with the British deaf community and for creative purposes, e.g., for storytelling. Associating BSL and SSE with specific groups of people and/or in specific domains in these ways are examples of iconisation.

SSE was described as an ‘oral approach’ (MC07), and associated with mainstream oral schools (i.e., schools for hearing children). Participants described how deaf adults who went to mainstream schools (ex-mainstreamers) tend to use SSE and that SSE is a product of mainstream education (e.g. 13). GW09 described young mainstreamers as being ‘forced’ to use and interact in SSE. Recent reports show that teachers and/or other professionals working with deaf children in mainstream schools who tend to be hearing are unlikely to be fluent in BSL thus may use signing that is heavily influenced by English (CRIDE 2023; DfE 2023). This would inevitably affect the variety of signing acquired and produced by the children. In addition, ex-mainstreamers may also not be aware that the signing that they use is different from BSL.

GW23 discussed how they attended one deaf school in Scotland where “everyone deaf used BSL” and then later moved to a hearing school where “the signing was different”, where BSL and English were mixed to a greater extent. Other participants (14) associated their SSE with their educational experience but also described how they began to use BSL after leaving school or ceased using SSE once they had left.

This was not always the case for everyone. Some commented that people continued to use SSE after leaving school because they hadn’t been taught BSL or had not been exposed to BSL in the deaf community or came from hearing families (15). Consequently, they continued to use SSE ‘out of habit’ (LN03),

As well as associating SSE with mainstream schools using oralist approaches, SSE was discussed in context of oralism more generally (e.g. 16). Participants described BSL as being “more for the deaf” while SSE “is more for the oral or the hearing” (BF16).

In addition, the use of SSE was associated with hearing people (e.g. 17). BM09 described how they ‘changed’ to SSE when they married a hearing person or would ‘sign and speak’ with hearing family members. GW24 described how they used SSE with a hearing person, or a hearing person tended to ‘be SSE’ because they ‘speak and sign’ (GW10). GW25 described how hearing-parented families may use SSE but deaf-parented families would use BSL signing.

In contrast to SSE, BSL was described as a language used in the British deaf community (e.g. 18). It was a part of ‘deaf culture’ (BM10), people used it when they were with the ‘deaf community’ (BM09) and something that ‘deaf people know and do’ (BM08). Whilst SSE was associated with educational settings and the teaching of English, MC06 described BSL as ‘our sign language’ and ‘the national sign language of the deaf community’.

7.3 Evaluation/preferences

The next theme frequently provided value judgments on BSL/SSE, which Irvine and Gal (2000) attribute to iconisation. As mentioned earlier, Irvine and Gal (2000) explain that some dialects or accents are deemed more esteemed than others, while others are seen as non-conventional or stigmatised. Generally, respondents expressed more positive opinions, with many stating their preference for BSL over SSE (as in (19)), which suggests participants had higher regard for BSL compared to SSE. One said that they understood BSL-using interpreters much better. BSL was described as ‘natural’ (BF23), ‘easy to understand’ (BF07), and ‘perfect’ (MC06) – best for explaining everything.

In contrast, opinions about SSE were more negative and users of SSE were viewed more negatively compared to those who used BSL. SSE was could ‘put you to sleep’ (BF05), ‘awkward’ (BM06) and ‘flat’ (GW01). There was ‘more effort involved’ (BF17) because ‘the signing takes longer and drags on a bit’ (GW01). Many said that they preferred BSL, feeling comfortable in it as a ‘natural language’ whereas SSE was considered to be ‘artificial’ (BL17) and ‘robotic’ (BM21). The value judgements of BSL and SSE here exemplify fractal recursivity, evident as BSL and SSE are positioned at opposite ends, positive versus negative.

BSL fluency was discussed – i.e. it can be ‘signed fast and fluently’ (BL22), as a language that ‘flows’ and more ‘animated’ (BM07). Its informal, casual nature meant that respondents felt conversing in BSL was easier and more effective therefore smooth and uninterrupted, unlike SSE (20).

Some discussed whether they understood other people using BSL or SSE (e.g. 21). Respondents said that BSL was ‘easy to understand’, partly based on facial expressions used (BF07). Others signed that interactions in BSL with other deaf people were ‘more funny, quick-witted and sharp’ (GW01).

In contrast, some respondents, particularly older signers, said that they could not understand SSE (e.g. 22). MC06 said that when speech was added to BSL, it was difficult to follow.

Three respondents did frame SSE as being useful for encouraging good literacy in English and with spelling (BF06). BF10 was grateful to have SSE to enable access to English, ‘I have to say thanks to SSE for allowing me to learn [English]’. One mentioned that BSL was used to educate him, resulting in him having a poor education, indicating negative language attitudes for BSL. BL32 said they used both, depending on who they were with, ‘adjusting to match how they sign’.

7.4 Awareness of language status

Language awareness here refers to when participants are aware of signing style differences and are conscious of what style they adopt themselves. Many felt that BSL and SSE are completely distinct signing styles. BM24 said, ‘it’s definitely different because BSL uses much stronger facial expressions, is denser and has its own slang’.

As noted in Section 2.1, prior to the 1970s, BSL was referred to as ‘signing’ or ‘deaf sign’ before it was known as British Sign Language (BSL). This subject came up a few times with BL10 mentioning this in the way sign language was labelled, ‘it wasn’t known as BSL then, it was ‘deaf people’s signing’, and that was until people started talking about BSL and SSE’. LN29 reported that there are older deaf people still saying they are not BSL signers, not understanding the label. They view ’BSL’ as a new language (23).

Some older people commented that they do not use BSL as it is something used by younger people (as noted in 24–25 from signers aged 60 and above).

Older signers in the BSL Corpus (typically 65+) reject the label ‘BSL’ as a descriptor of their signing and view BSL as something new, something that younger people do. BL10 in their 40s said that ‘those in my age group and older from my circle’ are BSL signers implying that younger signers used SSE. Both are examples of iconisation, as BSL and SSE are attributed to people of specific age groups.

Some people said they were not sure what SSE was, having ‘never heard about SSE’ (BM26) or were previously unaware of SSE, thinking they were using BSL. As discussed earlier, the transition between their realisation about using SSE and the time when they began to use BSL overlaps with leaving oral education and joining the deaf community (as in 26 from a signer in their 60s).

Some reported learning BSL linguistics in their BSL teacher training (27 & 28) from signers in the 41–65 age group.

7.5 Being aware of the continuum

In our review, we described how varieties of sign language use can often be considered as occurring on a continuum. LN25 explained ‘there is a range of forms which goes from very strong BSL, to somewhere in the middle, along to SSE at the other end’. Some descriptions of SSE in the data, however, do point to some awareness amongst participants of this continuum. BF01 explicitly stated that although they used the term SSE all the time, they wondered if it was the right term because ‘one person using SSE and another person using SSE are still different from each other’ and that ‘you can’t say that SSE is this since it depends on the person’. BL08 thought SSE was a part of BSL and ‘it sits under the umbrella of BSL’. BL08 again described SSE as ‘a clash between two languages (English and BSL)’ which suggests some awareness of SSE being a result of language contact (although framed in a negative way). BM16 stated sometimes they ‘think people who are signing SSE have BSL in there too’.

Based on the structural descriptions above, while some deaf signers do recognise a continuum, it appears that BSL and SSE are for the most part defined by referring to extremes i.e., sign order in BSL is described with reference to specific examples in non-English-like order (e.g. question signs are placed at the end) while SSE follows English. This is an example of fractal recursivity.

8.1 Language differentiation in BSL teaching and sign linguistics

Irvine and Gal’s (2000) semiotic processes can be clearly seen in BSL curricula and teaching materials. Sign language textbooks/sign language instruction books (e.g., City Lit 2008) often describe BSL with reference to specific structures (e.g., emphasis on topic comment structures, wh-final elements, and non-English sign orders), despite their low frequency (Hodge et al., in press). The reverse of this is also seen i.e., where English-like structures are downplayed (Fenlon 2019). For example, the Signature BSL Level 2 assessment criteria require use of “BSL sign order”, where full points are given when the “candidate is able to use BSL sign order, with no English structure, most of the time” (Signature 2022). The focus on infrequent constructions, the least similar to English, and the lack of acknowledgement of frequent constructions in BSL that actually are English-like points to erasure. Some aspects are also fractal recursivity, as BSL and English are described on extreme ends of the continuum. There is no mention of language mixing in curricula, nor is this allowed in assessments.

It is interesting to consider whether the ideology that BSL is totally different from English exists in the deaf community and this is reflected in the BSL teaching curriculum, or vice versa. Either way, this ideology may well have its roots in the early linguistic recognition of sign languages as legitimate languages. The push by linguists to show that BSL is not simply “English on the hands” may well have resulted in the fractal recursivity that we see between BSL and SSE. This could be compared to the ongoing debate within sign linguistics about the role of gesture in sign language (e.g. Quer 2020; Schembri et al. 2018). That is, the push to show that sign languages are not the same as gesture used by non-signers has led many linguists to reject outright the possibility that any part of what signers do might be gesture or gestural. These academic ideologies are sometimes reflected in the deaf community but sometimes resisted or transformed (Kusters and Sahasrabudhe 2018).

8.2 Language mixing as translanguaging

In terms of what SSE actually is: a translanguaging framework provides a useful explanation. This framework views language practices not as separate, bounded systems but as dynamic and fluid. Within this perspective, signers use various resources from both BSL and English when they communicate from whatever they have available in their linguistic repertoire, including signs, words, gestures, writing etc. Lines between what gets called BSL and what gets called English may well be drawn partly on modality, but what sits in between also involves social/political issues and power dynamics. Because of this, when the visual modality is involved, and questions arise about whether what is produced is BSL or SSE – that cannot be determined on a structural basis alone. As noted by Sutton-Spence and Woll (1999), it does in large part come down to what the signer believes they are producing. We conclude that it is difficult to distinguish between BSL and English-based signing varieties. We suggest that those varieties are better understood from a translanguaging perspective where the amount of English influence on BSL utterances is largely dependent on the signers’ knowledge of BSL and English, as well as context. It can also depend on how signers perceive each other’s language skills. For example, if each person knows that the other can understand a wide range of signing varieties, they may use any of the semiotic resources they have. But if one knows that the other person does not know many varieties, then they will accommodate accordingly (Moriarty and Kusters 2021). Translanguaging practices like these in deaf communities have been widely discussed, particularly by deaf scholars, with De Meulder et al. (2019) explaining that deaf people do engage in translanguaging practices, and they can be quite creative in using communicative strategies – therefore “translanguaging is central to the deaf experience” (p. 11).

Translanguaging is a recognition of the communication practices adopted by bi/multilingual people, including deaf signers, to achieve mutual understanding. However, some are concerned that ‘the ‘translanguaging turn’ (García and Li 2014) challenges the six-decade long project of sign language linguistics and by extension deaf studies to legitimize the status of sign languages and, correspondingly, the right of deaf people to acquire and use a sign language in educational and community settings’ (De Meulder et al. 2019: 2). Deaf scholars argue that translanguaging practices in deaf communities should be recognised and valued. However deaf children still have a right to be exposed to fluent models of natural sign languages, providing them with much needed access to language to develop ways to interact with the world around them (Henner and Robinson 2023).