Plenary Speaker Mark Sebba (Lancaster University)

Mark Sebba is Reader in Sociolinguistics and Language Contact at Lancaster University. He has done research on pidgin and creole languages and on conversational code switching in bilingual communities. Through this he developed an interest in the Sociolinguistics of Orthography, a relatively unexplored field which examines the cultural and social aspects of spelling and writing systems. Spelling and Society: The culture and politics of orthography around the world was published by Cambridge University Press in 2007, and a collection co-edited with Alexandra Jaffe, Jannis Androutsopoulos and Sally Johnson, Orthography as Social Action: Scripts, Spelling, Identity and Power was published by Mouton De Gruyter in 2012.

His most recent interest is in written bilingual and multilingual texts – magazines, websites, emails and other texts which contain a mixture of languages. His approach includes what has traditionally been called code-switching (i.e. the use of two languages within one text) but goes beyond that, to explore multilingual texts as literacy practices which draw on different repertoires of languages, visual images, spatial arrangements etc. A collection of papers co-edited with Shahrzad Mahootian and Carla Jonsson, Language mixing and code-switching in writing: approaches to mixed-language written discourse, was published by Routledge in 2012.

He has also recently been doing research on the way languages and language communities are enumerated, and has a number of papers in the pipeline regarding the 2011 United Kingdom census, which was the first to ask language questions in England.

Spelling Sylheti: phonology and practice

Mark Sebba, Lancaster University

Conventional wisdom within linguistics for most of the 20th Century was that orthography was the science of ‘reducing languages to writing’, in other words to create a phonemic script on structuralist principles, to allow a language to be written down and read. In fact, orthography was studied almost exclusively from this point of view – as a kind of encoding and decoding technology which was a prerequisite for ‘literacy’.

An alternative viewpoint would be a practice account of orthography, which sees orthography (or spelling) as a set of practices which are part of people’s everyday communication. These may be formal and relatively invariant practices (such as spelling in strict conformity with dictionary rules, as required at school) or informal and often variable practices, such as the now familiar abbreviated spellings used in text messaging. Such practices are potentially heterodox – for example, they may draw on conventions from more than one language and can incorporate popular ideas about the use and symbolism of particular letters.

In the context of a language where a traditional script or orthography is not available, or is unlikely to be learned, users of the language may develop a set of practices which work for their purposes, but are far from what the phonemicists would recommend.

In this talk I will propose that we should view orthography as embedded within practices of reading and writing, within a broader sociocultural account of people’s everyday literacy. This fits with a view of literacy which has been called ideological by Brian Street (1983), although I would prefer to call this a sociocultural model. Within such a model, I argue for a practice approach to orthography, in the hope that it also provides a practical way forward for literacy, without the need to ‘reduce a language to writing’.

Street, Brian 1984. Literacy in Theory and practice. Cambridge: Cambridge University Press.


“Mapping of Spirantization and De-aspiration in Sylheti” 

Satarupa Sen (The English and Foreign Languages University, Hyderabad, India)

Sylheti is an Indo European language spoken throughout Sylhet division, Bangladesh and is the dominant language of Barak valley in southern Assam, India.

The spirantization in Sylheti identifies with the same occurring in the neighbouring language Assamese. In Assamese, aspirated stops /pʰ/ and /bʰ/ are spirantized as /f/ and /v/ respectively in word final position, blocking it in word initial and medial positions and also, only the labial stops undergo spirantization at the word final position unlike coronal and velar stops.

But in case of Sylheti, Spirantization is found in abundance irrespective of the position of the sound in a word. For example,

Bangla                                                   Sylheti                                           Gloss
/kore/                                                     /χɔria/                                            Having done
/palak/                                                   /falaχ/                                            Spinach

Again, deaspiration process in Sylheti is more of a common phenomenon irrespective of the place in a word. Eg,

Bangla                                                   Sylheti                                             Gloss
bʰai                                                        bai                                                    brother
ɒbʰinɔnd̯ ɔn                                         ɒbinɔnd̯ ɔn                                     congratulations
b̃ad̯ ʰ                                                      band̯                                                 dam

In Assamese, coda is de aspirated when followed by an aspirated onset. /f/ and /v/ change to /pʰ/ and /bʰ/ respectively in coda position without losing their [+asp] feature. Aspirated stops in onset position never undergoes the feature [+asp] alteration. But in case of Sylheti, the aspiration in plosive is lost in word as in /ɑːdʰ/ (half) + /χɑn/ (dempstrative) = /aːdχɑn/ (the half portion).

This paper aims to investigate and draw a pattern of Fricativization and De- aspiration process in Sylheti which form the idiosyncrasy of the language and sets it apart from both Standard Colloquial Bangla and Assamese.

Keywords: Sylheti, Assamese, Phonology, Spirantization, De-aspiration

Mayenin, M. Larakoron My Sylheti Grammar.Lulu.com.
Tuṅga, S. Ś. (1995). Bengali and other related dialects of South Assam. Mittal Publications.

“Does Sylheti have consonant clusters?”

Elizabeth Eden (University College London)

Sylheti-English sequential bilingual children perform better with English word-initial clusters which are found with a relatively high frequency in Standard Bengali (McCarthy, in preparation), implying a transfer effect from their L1.

However, by examining the distribution of consonantal segments word-medially in Camden Sylheti, as documented in the SOAS Sylheti Dictionary, I will show that the native lexicon has only (C)V(C) syllables. I will briefly discuss repair strategies observed in the Sylheti lexicon for historic loan words, and question which of these are still extant. I will also demonstrate the computational tools which I have developed to help with such analyses.

A simple view of syllable structure is that any sequence of consonants that occurs word-initially is a valid syllable onset, and any sequence that occurs word-finally is a valid coda (Zec 2007). However, a closer cross-linguistic examination shows that consonant distributions may vary between word-edge and word-medial positions (Kaye 1990).

Word-edge consonant sequences in Sylheti are relatively rare. Word initial clusters are predominantly found in loan words e.g. English [pɾofɛsaɾ], [skʊl] and Sanskrit [gram], [pɾaʃnɔ]. Likewise, most of the final consonant sequences in the Dictionary are either loan items e.g. [ɡɪfʈ], or morpheme edges which do not surface without a following vowel e.g. [afn-]. However, word-medial consonant sequences are relatively common in the native lexicon.

Word-medial sequences can be divided into valid word-final sequences and remaining consonants. In English, for example, [ŋɡɹ] in angry splits into [ŋ]/[ɡɹ], implying that [ɡɹ] is a valid word-medial complex onset. For Sylheti, I demonstrate that the word initial “complex onsets” are not in fact found word-medially; they never follow a valid word-final sequence. I similarly demonstrate only 2 of the 27 word-final sequences are found as internal codas. The status of these 2 sequences (as allophonic variants or loan phonology) is debatable. By contrast, all word-initial singletons are also found as word-medial onsets, and all word-final singletons are also found as word-medial codas.

I conclude that all word-medial consonant sequences in the native lexicon contain syllable boundaries. Two thirds of these sequences obey the Syllable Contact principle, i.e. drop in sonority across the boundary. However, Sylheti does also permit heterosyllabic sequences of rising sonority, e.g. [kɾ], in order that the more important constraint on complex onsets is obeyed (Clements 2009). These heterosyllablic sequences appear to assist sequential bilinguals in acquiring their tautosyllabic English counterparts, even though the English structures are not licit in Sylheti.

“Scoping out negation: a lexical functional account of negation in Sylheti”

Zander Zambas

Scoping out Negation

“Differential Object Marking in Sylheti”

presentation-Rob-DSCN9756Robert Laub (SOAS, University of London)

The Sylheti language exhibits use of Differential Object Marking (DOM). This means that an object may or may not be marked with a particular suffix. There are a number of factors that lead to the presence of absence of this marking. DOM is a common feature of languages throughout the world, including those within the same language group. My research looked at DOM within the context of Sylheti and other IndoAryan languages. I also looked at what factors contributed to the use of the object marking in Sylheti.

Sylheti marks its case with suffixes. Nominative uses the suffix e, genitive as or or r, and a locative t. This study focuses on the suffix re, which can be found on objects.

Aissen (2012) laid out hierarchies showing where a noun phrase is more likely to be overtly marked. I have used the following hierarchies in my investigation:

Animacy: Human > Animate > Inanimate

Definiteness: Personal pronoun > Proper name > Definite NP > Indefinite specific NP > Nonspecific NP

While some languages only use one of these two hierarchies, IndoAryan languages use both. Another motivation for using morphology to express definiteness is that IndoAryan languages do not use articles to convey definiteness.

I found three ways that Sylheti can show definiteness of NPs: standalone nouns, nouns preceded by a number with a classifier suffix, or a noun with the classifier suffix.

My examples of Sylheti come from a consultant who speaks Sylheti and English. I investigated his Sylheti with English prompts, asking him to translate them for me. Although this method is not perfect, it provides a small insight into how DOM in Sylheti works.

In looking at the words on the animacy scale, I found instances of object marking on all three categories. However, they are more likely to have the object suffix the higher they were on the animacy scale. On the definiteness scale I divided up standalone nouns into names, singulars, and plurals/mass nouns. Giving this a closer look shows that names are more likely to have the object suffix, plurals and mass nouns are less likely, and singulars are in a middle ground. Those in a middle ground in both scales have the most variation with the suffix, and it seems that this is used to express specificity.

Aissen, Judith (2002). Differential Object Marking: Iconicity vs Economy. Natural Language & Linguistic Theory

“Relative Clauses in Sylheti”

Caoife Garvey (SOAS, University of London)

Relative Clauses in Sylheti

“A Cross-Linguistic Perspective on Converb Constructions in Sylheti”

presentation-Sarah01-Photo 12-05-2016, 18 17 27Sarah Dopierala (SOAS, University of London)

There are many terms that refer to nonfinite verb forms marking adverbial subordination including: gerund, adverbial participle, and converb. (Haspelmath 1995) Although it is difficult, or perhaps impossible, to have cross-linguistically valid categories, converbs are one such category whose characteristics can be observed across languages. (Haspelmath 1995)

The purpose of this presentation will be to identify those characteristics that are typical of converbs cross-linguistically, as described by Haspelmath (1995), and examine how certain Sylheti constructions exhibit these characteristics. Using examples from data collected during elicitation sessions with a native Sylheti speaker in the Field Methods course at SOAS 2015-2016, I will argue that Sylheti does seem to have converb constructions.

In the presentation, I will focus on three aspects of Haspelmath’s (1995) converbs 1) the absence of tense, aspect, mood (TAM) or agreement marking 2) the subject of a converb being shared with the subject of the matrix clause 3) and the reduplication of the converbal form.

The first aspect I will examine is the absence of TAM and agreement marking on verbs ending in -ja (also -ia). The -ja form never occurs in a sentence without a fully inflected verb form; its form does not change based on differences in person or number; and its form does not change based on the tense of the fully inflected verb.

The second characteristic I will describe is the subject of a converb usually being coreferent with the subject of a superordinate clause. This means that the subject of a converb is often omitted. Sylheti seems exhibit implicit-subject tendencies as the subject of -ja is not always overtly stated. (Haspelmath 1995: 9-11)

  1. [ɸua duri -ja            gi  -ja           ex   balʈi         ɸani     ani      -ja

boy run   -CONN go  -CONN  one bucket       water   bring  -CONN

zanala     mukai    di        -s            -e]

window   near       give    -PFV      -3P.IF

‘The boy goes running, brings a bucket of water, and pours it through the window’
(Dopierala 15/1/2016)

Finally, I will show that the -ja form can often be fully reduplicated and that this reduplication usually performs some sort of adverbial-like function.

Through discussion of these and other relevant characteristics I hope to show the existence of converb constructions in Sylheti, and also implicitly support Haspelmath’s (1995) claim that CONVERB is a category that appears across languages.

Haspelmath, M., 1995. The Converb as a Cross-Linguistically Valid Category, in: Converbs in Cross-Linguistic Perspective: Structure and Meaning of Adverbial Verb Forms – Adverbial Participles, Gerunds -. Mouton de Gruyter, Berlin.

“Additive focus in Sylheti”

Heather Brown (SOAS, University of London)

This presentation will be based on research conducted for my masters dissertation examining the syntactic status and information structural role of the Sylheti particle. Cross-linguistically a number of researchers have suggested that a class of focus particles can be observed: particles which have an information structural role in the sentence, conferring focus on the elements within their scope. I will ask whether the Sylheti morpheme is a member of this group, and how the constraints on its placement can be modelled.

In terms of meaning, I will argue that adds the element within its scope to some previously established set:

  1. Context: She bought vegetables from the shop

gʊst=ɔ                         kɪn-tʃ-ɔɪn                     dukan taki

meat=AD.FOC           buy-PST-3.FML          shop     from

“…(she) also bought [meat]F from the shop”

As such, the particle fits into the category of an additive focus particle as discussed by researchers such as König (1991) and Dik (1997). In this it is similar to the Bengali form =o. While Sylheti and Bengali are closely related, it is notable that, although Bengali has been argued to show discourse configurationality (eg. Butt and King 1998), limited evidence of this was seen in the Sylheti data collected for this project.

I also examine the placement of the form and its syntactic status, suggesting that it is best considered a clitic. Its status as a discourse clitic rather than a case clitic is demonstrated both in its meaning and its distribution: it shows considerable freedom of placement, and is able to mark both focus on a single argument or on a larger portion of the sentence. A key constraint on its placement is noted in that it is unable to appear utterance-finally, and I argue that this is best modelled through a series of ordered constraints.

Overall I will argue that the meaning contributed by this particle can be analysed similarly to additive focus forms in a variety of languages, but that its morphosyntactic status and positioning present a greater challenge, particularly in their divergence from similar forms in related languages.

Butt, M. and King, T. H. 1998. Interfacing phonology with LFG. In: Butt, M. and King, T. H. On-line Proceedings of the LFG98 conference. Stanford, CA: CSLI Publications.
Dik, S. 1997. The Theory of Functional Grammar Part 1: Structure of the Clause. Berlin: Mouton de Gruyter.
König, E. 1991. The meaning of focus particles. London: Routledge.

“The role of Sylheti in the investigation of the linguistic and cultural mediation of disaster and health messages about heatwaves and cold spells within the Bangladeshi community in Tower Hamlets”

Chris Tang (King’s College London) & Lubaba Nusrat Khalil (King’s College London)

This talk reports on a project looking at the role of language, culture, ethnicity and community in the dissemination of disaster and health messages about heatwaves and cold spells to the Sylheti speaking community in the London Borough of Tower Hamlets. One of the most frequently employed strategies in public health communication with Sylheti speakers with lower levels of English proficiency is to translate into Bengali. This approach, however, fails to take into account Sylheti’s lack of a written form, the diglossic relationship between standard Bengali and Sylheti, and the need for cultural as well as linguistic translation. Drawing upon theory from the fields of risk communication, public health science and cognitive linguistics, the project explores the potential for recruiting young members of the Bangladeshi community as cultural and linguistic mediators of key information prior to or during a heatwave or cold spell.

The talk will focus on two key points. Firstly, we report on data from a series of focus groups held with younger and older members of the Bangladeshi community in East London, in which there was a primary focus on participant engagement with warning and health advice about extreme temperatures typically found in news reports or in government outreach. The different ways in which younger and older members of the community responded to, interpreted and appraised this information will be discussed in relation to the role of Sylheti in the cultural and linguistic mediation of key messages. Secondly, using this empirical study as an example, we report on the challenges in obtaining qualitative data in research involving Sylheti speakers, in particular in relation to design of oral research materials and the transliteration and translation of audio data. 45

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s