Literacy in Contact and in Context : Multilingual Reading and Writing along the Silk Routes

Received on: Apr.1st, 2020 Accepted on: Sep. 03rd, 2020 Abstract: According to UNESCO, at least 2500 languages are vulnerable. Chinese, English, Spanish, Arabic, Hindi, Portuguese, Bengali, Russian, Japanese, French are “hegemons” each having at least 100 million native speakers and accounting for over 51 percent of the global population. Half of the hegemons are written with an alphabet. For the non-alphabetic group, native speakers may read and write in logographic (e.g. Chinese) or syllabic writing systems (e.g. Devanagari) or both (e.g. Japanese). In languages that are spoken by less than one million people, Latin, Arabic and Chinese writing systems dominate but they do not always map to local dialects transparently. Multi-literacy is a growing global phenomenon particularly in Asia. In the 21st century, access to electronic literacy will include multi-literate speakers. However, multi-literacy brings questions. Multi-literacy is as old as civilization due to spoken language contact in commerce, ideology and religion. Literacy adapts to new technology via codification of symbols allowing multi-literacy to grow. Documentation of writing has a history but it is not prominent in global policy making. Programmes to develop literacy are reserved for monolingual ‘hegascripts’ (dominant languages) e.g. English. However, neglecting diversity in writing systems in developing countries risks more inequalities if indigenous language speakers are taught literacy in their non-native language only.


Introduction
A person who habitually speaks more than one language can be called a bilingual or multilingual whereas a person who writes in more than one script is bi-literate or multi-literate. A Google search returns more than 100+ million hits for bilingual but only 2,500 for bi-literate illustrating the lack of 3 See: https://contentmarketinginstitute.com/2011/08/content-marketing-language-barrier/ understanding of the diversity of writing systems globally. This is becoming an issue as technological advances (electronic, social media, texting) demand multi-literacy. Indeed, there has been a rapid increase in written languages used online: Arabic (2500+%), Russian (1825+%) and Chinese (1478+%) among others (see Figure 1). Multiliteracy is important for digital citizenship globally but this is most acute in officially polyglot areas including China, Hong Kong and India. One goal of this paper is to introduce a conceptual framework to motivate research into the effects of multiliteracy on neurocognitive processing and ergo on healthy aging and levels of educational attainment in multiliterate environments. What is most intriguing about multi-literacy are questions about how our neural systems have adapted the mechanisms used to process more than one language (each with their own idiosyncrasies) in the brain (CHEE et al., 1999;WEEKES, 2005; and whether this enhances cognitive reserve that is vital for healthy aging (ABUTALEBI et al. 2014(ABUTALEBI et al. , 2015a(ABUTALEBI et al. , 2015b(ABUTALEBI et al. , 2016 as well as educational policy (ASFAHA e KROON, 2011). Another goal of the paper is a re-conceptualisation of multi-literacy from a neurocognitive perspective in order to distinguish between bi-scriptal (different scripts learned) and di-scriptal (same script learned) bilingual speakers (who use different languages) and monolinguals who speak the same language and thus read the same script but may be bi-scriptal within their own language e.g. in Japanese. This new conceptual approach is necessary because multiscript literacy is determined by at least two constraints: the type of language (same family versus different family) and type of script (same script versus different script). A novel feature of the framework is the distinction between writing systems that are similar but not equivalent e.g. simplified and traditional characters in Chinese languages and Bokmål and Nynorsk in Norwegian languages -and how they represent their own language families. For example, Bokmål and Nynorsk di-scripts use Latin alphabets to transcribe a common language and both are written phonetically. However, logographic characters used to write Chinese dialects are non-alphabetic and thus cannot be used to write different languages below the level of the syllable. This creates a barrier for adopting logographic scripts to write in a non-related language e.g. Chinese characters used to write Mongolian and Tibetan but not necessarily an obstacle e.g. Japanese and Korean use Chinese logographs. The key conclusion for policy purposes is that instruction in a non-native script as the medium of instruction creates difficulty for learning to read and write if their native language at home (or school) is taught in an opaque script e.g. learning to read and write Chinese characters when the home dialect is Hakka or Hokkien or even a language like Nepali.

Background
The focus of research on literacy is Indo-European languages that use an alphabetic script (e.g. COLTHEART et al, 2001;PLAUT et al., 1996;SEIDENBERG e MCCLELLAND, 1989;ZORZI;HOU-GHTON;BUTTERWORTH, 1999). Such models can explain literacy in English, German (ZIEGLER; PERRY; COLTHEART, 2000), Italian (ZORZI, 2010) and Russian (ULICHEVA et al., 2015). Some models can also explain acquisition of literacy in monolingual speakers learning alphabets (DAVIES e WEEKES, 2005;CASTLES, 2007) and loss of literacy following brain damage (HRICOVÁ e WEEKES, 2012;SENAHA e PARENTE, 2012;SENAHA et al., 2006;WEEKES, 2007;WEEKES et al., 2012;WEEKES, et al., 2013). However, it is not clear if models extend to monolingual speakers learning non-alphabetic scripts (GUAN & WEEKES, 2019) and losing literacy through brain damage or even healthy aging (LEUNG, et al., 2012;WEEKES, 2000WEEKES, , 2005WEEKES, , 2012YIN, 1997;WE-EKES;LIN, 1998;WEEKES e CHEN, 1999;WEEKES et al., 2006a;WEEKES et al., 2006b;YIN e WEEKES, 2003;HE;WEEKES, 2005). Equally, such models are not well equipped to explain the acquisition and loss of literacy in bilingual mono-scriptal speakers ( VAN HEUVEN et al., 2018;WEEKES et al., 2013;RAMAN e WEEKES, 2003;2005a;2005b;WILSON et al., 2007)  Kroll and colleagues assume separate lexica for L1 and L2 word forms that are connected via asymmetrical connections shaped according to the amount of exposure in each language with each set of word forms connected to a common semantic system (though see KROLL e DIJKSTRA 2002) that allows for different writing systems to have differential effects on reading (HOSHINO et al., 2008). These models of bilingual reading can be considered language specific i.e. lexica represent independent orthographic word forms that are specific to one spoken language. By contrast, language independent models of bilingual written word recognition assume that written and spoken word forms and their conceptual properties shared a common network for all languages (VAN HEUVEN et al., 2018). Furthermore, written word naming was better than spoken word naming in both languages showing that -although there was a task effect -linguistic dissociations were not due to task difficulty as that variable would be expected to have an interactive effect. It is notable that Greek Cypriot speakers use the same script as other Greek speakers, but they pronounce letters according to the Cypriot dialect. This distinguishes them as di-scriptal not bi-scriptal (like Japanese speakers).
It has been argued that Greek Cypriots (who may be called bi-dialectal speakers) have a cognitive advantage (ANTONIOU et al., 2015). However, it is not certain that spoken language (code) switching is the reason for this advantage. An alternative hypothesis is that, because Cypriot

7/17
alphabetic writing with embellishments for tone in Mainland China (called pinyin). Multi-scriptalism is thus a poly-dimensional construct that can be conceptualised using dimensions that are intralingual (one language) and interlingual (two languages).
A framework for multi-scriptalism presented in Figure Three can be used to conceptualise script diversity according to spoken language similarities and for testing hypotheses about literacy in monolingual, bilingual and multilingual speakers (WEEKES, 2005(WEEKES, , 2012.  Letrônica, Porto Alegre, v. 13, n. 4, p. 1-17, out.-dez. 2020 | e-37538   is a principle that is guiding these initiatives but multi-literacy is much neglected (NAG et al., 2012). Indeed, for monolingual speakers who use a 'dialect' or speak a non-dominant language at home, multi-literacy has been demonised as a threat to learning (undermines scholastic achievement/EFA goals) by governments and researchers globally. Teaching multiple scripts is considered negative in mainstream education.

8/17
International agencies e.g. UNESCO must redefine multi-literacy as an opportunity not a threat. This is timely given that literacy is an economic and geopolitical imperative and multiliteracy adding resource to human capital with interesting consequences for contact along OBOR ( Figure Seven). One contention is that multi-literacy may deliver advantages beyond educational achievement in terms of preserving cognitive reserve that is vital for healthy aging. There is no argument that these benefits must be investigated given that multi-literacy -in the digital age -is essential for commerce, education, politics and trade most especially for Chinese seniors who are recognised as early adopters of technology. Results of these investigations contribute to EFA goals and UNESCO priorities. My recommendation is to report examples of cross-script comparisons in acquired dyslexia and dysgraphia (WEEKES, 1996;2019;WILSON et al., 2012;Figures Three to Six).

Routes and Roads
There is scope for further research given widespread language contact along the Old Silk routes and the diversity of writing systems in OBOR. We have identified cases of multi-scriptal dyslexia and dysgraphia in China and have broadened catchment to cases in OBOR locations and studied the brain mechanisms for multi-literatism in bi-scriptal Chinese-English speakers (CHEE et al., 1999; as well as in acquired communication disorders in several languages. Green's (1986) Inhibitory Control (IC) model offers one motivation to study bi-scriptal speakers as a test of the language control hypothesis (ABUTALEBI e GREEN, 2007). Briefly, The language control system is assumed necessary for selection of lexical representations in bilingual discourse and thus potentially explains literacy in more than one language. However, literacy is more than just word retrieval and word production. For multi-literate individuals, literacy can involve switching between oral language and writing systems i.e. orthographies and scripts in everyday life -particularly in the educational context.