next day
all days

View: session overviewtalk overview

09:30-10:20 Session 1: Plenary Session I
Location: New York
Homo e-loquens

ABSTRACT. Nature loves variety. Unfortunately, society hates it. Milton Diamond

Human beings speak and interact in an impressive number of different languages varieties, and, within these varieties, in very many ways. This huge scale of variation between and within languages seems to be a more pregnant human characteristic (the homo loquens) than being smart (the homo sapiens). The variability of language has led to the formulation of the principle of inherent variation (cf. Weinreich, Labov, & Herzog 1968). Language variation can nevertheless be put aside to focus on the robustness or homogeneity of the internal, underlying grammar (Chomsky; i-language), but I want to argue that we need to address the pervasiveness of variability and changeability of language (Labov; e-language) in a fundamental way, to understand what language is and how language structure evolves.

The current 86 Indo-European languages can be traced back to a common ancestor of 8700 years ago (Gray & Atkinson 2014), which is only about 435 generations, and is an example of the variability and changeability of languages. I will discuss other, recent examples from the Dutch language area to illustrate the extent of variability. These examples relate to sounds, words and grammar.

Changeability and variability can be coped with by human adaptivity in communicative processes, in combination with impressive learning capacities. I will focus on adaptivity processes, also in relation to form-meaning correspondences and semantics. I will argue that human meaning is by its very nature incomplete in interactive processes, in fact driving the changeability of form-meaning relationships.

My conclusion is that unifying, structuring forces in the language variability space come from interactive processes. External forces unify and structure language variation and language. This is amply demonstrated by urban studies on language variation, but also by dialect maps showing the convergent force of social factors.

10:20-10:45Coffee Break
10:45-12:25 Session 2A: Language Contact and Change
Location: Brussels
Obersaxen – A Highest-Alemannic Sprachinsel. Linguistic change as a consequence of dialect contact


The present paper deals with the effects of contemporary globalisation and mobility on the variety spoken in a previously isolated and peripheral Sprachinsel. I will show that in the case of Obersaxen, a Highest-Alemannic Sprachinsel in the midst of a Romansh-speaking area, situated in the most western part of Grisons, Switzerland, linguistic changes on the structural level are to be attributed more to the influence of other Swiss German dialects than to contact with the surrounding language, as common in most Sprachinseln. Furthermore, I will also bring evidence suggesting that certain specific aspects of Highest-Alemannic phonology, already present when people left the canton of Valais and migrated to Grisons in the 12th–13th century, have been preserved, thereby demonstrating, unusually, that the phonological is more stable than the morphological and (morpho-)syntactic levels. For this reason, the study will take a sociolinguistic-variationist viewpoint focusing on how increased tourism in the past decades as well as rapidly emerging bilingualism among Romansh-speakers have fostered structural changes as a result of enhanced dialect contact, while interestingly only having had a limited influence at the phonological level.

Based on a corpus comprising speech-data collected in non-structured sociolinguistic interviews from 24 subjects equally distributed among three age groups, five linguistic variables (two phonological and three [morpho-]syntactic) have been analysed quantitatively in a combined real- and apparent-time approach. As a result of the analysis of this dataset, it is possible to describe changes that have taken place since the first half of the 20th century
(cf. Brun 1918; Hotzenköcherle 1962–2003), as well as to make inferences about present-day linguistic developments in the community.  



Brun, L. (1918) Die Mundart von Obersaxen im Kanton Graubünden: Lautlehre und Flexion. Frauenfeld: Huber. (= Beiträge zur schweizerdeutschen Grammatik 11).

Hotzenköcherle, R. (ed.) (1962–2003) Sprachatlas der deutschen Schweiz (SDS). Bd. 1–9. Bern/Basel: Francke.

The social meaning of contact features in ‘new speakers’ of Basque


Work on ‘new speakers’ of Basque has shown that they do not enjoy the same authenticity as their ‘traditional’ speaker counterparts (Ortega et al., 2014, 2015; Urla et al., 2018).  However, there is a dearth of research focusing on the linguistic system of these speakers and an understanding as to how the social meaning behind their use of contact-phenomena contributes to their (in)-authentication processes as members of the Basque-speaking community. The present study examines the variable use and attitudes towards two linguistic phenomena, namely, Differential Object Marking (DOM) and ergative-case marking (or lack thereof), with the goal to determine the way contact features achieve social value.

I build on my work on the spontaneous speech of these two phenomena by examining 24 new speakers’ (and 24 ‘traditional speakers’) attitudes towards Basque DOM by means of a matched-guise experiment (MGE) and subsequent metapragmatic commentary. In the MGE, participants listened to 8 guises; half of which were in Batua (Standard Basque) and the other half in a ‘local’ variety. Each variety was manipulated for the presence or absence of DOM or ergative-case marking. For each guise, participants responded to 18 questions such as “How Basque is this person?” using a 1-7 Likert-Scale. Results from the MGE were analyzed using a number of mixed-effects models and correlations in R.

Results from the MGE show that Basque-DOM in the Batua guise indexes ‘defectiveness’, ‘Spanish-based barbarism’ and ‘inauthenticity’, but it is considered ‘authentic’ in the vernacular variety. Lack of ergative-case marking is strongly associated with ‘new speakers’ and ‘lack of competence’, but goes unnoticed in the vernacular guise. Additionally, Batua guises ‘correct’ ergative-marking are viewed as containing more ‘mistakes’ than its vernacular counterparts. Finally, there is a continuum of authenticity, variety and contact features: Pearson’s correlation shows that if a speaker is perceived using ‘bad’ Basque, it is also perceived as ‘less Basque’, mainly in the Batua guises.

These results corroborate recent findings on the enregisterment of Batua as a non-authentic variety. Here, I propose that the indexical meanings DOM and lack of ergative-case marking in the two varieties are the result of a complex interaction between three ideologies (mother tongue ideology, nationalist ideology and monoglot ideology) which can be better explained once we take into account the social history of the political praxis within the Basque revitalization project.



Ortega, A., Amorrortu, E., Goirigolzarri, J., Urla, J., & Uranga, B. (2014). New Basque speakers: Linguistic identity and legitimacy. Digithum, 16, 47-58.

Ortega, A., Urla, J., Amorrortu, E., Goirigolzarri, J., & Uranga, B. (2015). Linguistic identity among new speakers of Basque. International Journal of the Sociology of Language, 231, 85-105.

Urla, J., Amorrortu, E., Ortega, A., & Goirigolzarri, J. (2018). Basque Standardization and the New Speaker: Political Praxis and the Shifting Dynamics of Authority and Value. In Lane, P., Costa, J. & De Korne, H. (Eds.), Standardizing Minority Languages: Competing Ideologies of Authority and Authenticity in the Global Periphery. New York: Routledge.

Old and new forms of like in Cardiff - A Tale of two Valleys


When an innovative feature spreads rapidly into the linguistic system, it is rare that it does not interact with and affect older related forms in some way. The several forms of like in English are a case in point and in many English-speaking communities they are used alongside each other. An example of this can be seen in the extract below of Karim, a 27 year old from Cardiff, describing some of the students and teachers at his former school.

  1. They [= some students] could do half a class, but for like [DM] the second half they'll just have a spasm and like [DM] run out, like [CF]. Yeah, yeah, bouncing off the walls, like [CF]. So they'll [= the teachers] like [DM] mark them and they're like [QL], "yeah ok for the first half” and like [QL] “no, he messed up on the second half."

The frequency of quotative like (QL) and clause initial/internal discourse marker like (DM) have increased rapidly in the English speaking world since the end of the twentieth century (D’Arcy 2017, Durham et al. 2012, Tagliamonte and Hudson 1999). Both forms are associated predominantly with younger speakers (despite the fact that the initial innovators brought them into the system in the 1980s), whilst clause final like (CF) is an older form and one that is often associated with older, rural men (Miller 2009). In Wales, while the stereotype of the newer variables is that they come from the United States and ‘Valley Girls’, the association of the older form is with older men from the Welsh Valleys.  Using recent sociolinguistic interviews of men and women from three different age groups and older archival data of Cardiff speech to gain apparent and real time perspectives, this paper will examine how the over 5000 variants from the ‘different Valleys’ interact and how this has changed over time. Unsurprisingly only the two youngest generations have the quotative like variant, but all ages have the other two variants. The increase of the new variables has led to some restructuring of the older form, but not for all speaker groups: some of the young men use all three forms and have combined the old and new patterns. This may be tied to the continued association of the clause final form with Welsh (male) identity, while the other forms index youth.



D’Arcy, Alexandra. (2017). Discourse-Pragmatic Variation in Context: Eight-Hundred Years of  LIKE. Amsterdam: John Benjamins

Durham, Mercedes et al. (2012). Constant Linguistic Effects in the Diffusion of be like. Journal of English Linguistics 40(4) 316–337.

Miller, Jim. (2009). Like and other discourse markers. In Pam Peters, Peter Collins & Adam Smith (eds.), Comparative studies in Australian and New Zealand English, 317-337. Amsterdam: John Benjamins Publishing Company.

Tagliamonte, Sali A. & Rachel Hudson. (1999). Be like et al. beyond America: The quotative system in British and Canadian youth. Journal of Sociolinguistics 3. 147–172.

Variation within Frisian class II verbs: class-overriding productive past tense forms of the second person singular
PRESENTER: Anne Merkuur


This study evaluates recent developments in the inflection of Frisian verbs. Frisian has two classes of regular verbs (class I with an infinitive ending in - ə, class II ending in -jə), both with their own inflectional paradigms. The past tense form of class I verbs is similar to Dutch verbs. Class II verbs on the other hand show a different pattern with no dental suffix in the simple past:

(1)          Paradigms of class I and II           

                Person/Nr                         Class I                          Class II

PRS             1sg                                   bak-ø                         wurk-j-e

                    2sg                                   bak-st                        wurk-e-st

                    3sg                                   bak-t                          wurk-e-t

                    PL                                     bak-e                         wurk-j-e


PST             1sg                             bak-te[1]                wurk-e-ø

                    2sg                              bak-te-st                         wurk-e-st

                    3sg                              bak-te-ø                          wurk-e-ø

                    PL                                bak-te-n                          wurk-e-n


Several studies suggest that, under the influence of Dutch, class II verbs have a tendency to switch to the more Dutch like paradigm of class I verbs (Feitsma, 1982; De Haan, 1990; De Haan, 1997; Hoekstra, 1993). According to our dialect survey, however, these verbs do not switch to class I en masse. Furthermore, participants of the survey do inflect one paradigm form in a class I manner far more often than the other paradigm forms: The second person singular in the past tense. Do wurkest (you worked) thus for those participants becomes do wurktest (you worked) (2a):

(2)                                                                          norm                                                    alternative
a) Regular class II verb   Wurkje                 do wurk-e-st                                      do wurk-te-st
                                               work.inf              you work-pstclassII-2sg                you work-pstclassI-2sg

b) Strong/irr verb             Sjen                       do seach-st                                        do seach-de-st
                                               see.inf                 you see.pst-2sg                                you see- pstclassI-2sg


The same regular class I inflection is observed in Hoekstra (2013) on the past tense of 2sg forms of strong and irregular verbs (2b). We will argue that the class I inflection of 2sg forms spread from the strong/irregular verbs to the class II verbs, and is now the product of a class-overriding productive rule which states that the past tense of any 2sg form is formed by adding the past tense suffix –te/–de. Furthermore, we will show that this additional rule for 2sg past forms does not render the inflectional system with two regular classes unstable, thus accounting for the attested stability of class II as a whole.


Haan, G.J. de (1990). Grammatical borrowing and language change: the dutchification of Frisian. Journal of multilingual and multicultural development, 11, 101 - 118.

De Haan, G. J. (1997). Contact-induced changes in modern West Frisian. Us Wurk, 46(1-4), 61-89.

Feitsma, A. (1982). Oer it lienen fan wurden yn har gehiel en fon bûgingsütgongen apart'. Fryx, 3, 70 - 71.

Hoekstra, J. (1993). It feroaringsproses by de je-tiidwurden. Tiidskrift foar Fryske Taalkunde, 8, 43-51

Hoekstra, J. (2013). Do seachdest der idd aardig wasted út ja. De 2de persoan iental doetiid fan'e sterke tiidwurden yn it Frysk fan'e jongerein. Us Wurk62(1-2), 1-20.

[1] The choice for the –d or –t affix is a phonological issue; stem-final voiceless consonants take –t, elsewhere –d is inserted.

10:45-12:25 Session 2B: Interaction
Location: London
We are not listening to ourselves: Modelling listener responses as a discourse-organisational variable


There is ongoing debate about how best to define and model sociolinguistic variables, particularly discourse-pragmatic features. Interactional approaches have been used to define the variable context and the variants (Pichler, 2016). Simultaneously, interaction is increasingly being considered as a relevant conditioning factor in analyses of phonetic variation (Levon, 2016; Nilsson, 2015; Raymond, 2018).

In this paper I suggest that listener responses, or backchannels (Yngve, 1970), are firmly rooted in interaction and that we need to account for this in our definition of them as sociolinguistic variables. Previous analyses have quantified them relative to total or same-speaker number of words, or a set amount of time. Given that the listener responds to the other at any given point, contexts in which the variable can occur need to be defined based on the other’s linguistic production (similar to (Duncan & Niederehe, 1974)).

The central methodological proposal put forward is to calculate the frequency of listener responses on a turn-by-turn basis, as the number of listener responses by the recipient relative to the number of words in the ongoing (multi-unit-)turn produced by the interlocutor.

The analysis is based on an audio-recorded corpus of dyadic interactions between 16 Scotland-based participants (5m, 11f), totaling 17 hours of data. All conversations were transcribed in ELAN, listener responses were annotated and then coded for action type.

Zero-inflated Poisson regression models show that turn length is the most significant predictor for backchannel frequency. In terms of social factors, recipient gender is statistically significant, as is turn-holder gender: women listening to women backchannel at the highest frequency, and men listening to men at the lowest. Given the small and unbalanced sample, these findings can only be suggestive of a potential pattern and call for further research.

This analysis challenges the way we think about the sociolinguistic variable and variation as rooted in the individual: some variables are intrinsically interactional. It invites us to reconsider how we deal with Labov’s principle of accountability when analyzing variables beyond the level of the phoneme which might be rooted more deeply within the turn-by-turn unfolding of interaction than acknowledged in the past.



Duncan, S., & Niederehe, G. (1974). On signalling that it’s your turn to speak. Journal of Experimental Social Psychology, 10(3), 234–247. https://doi.org/http://dx.doi.org/10.1016/0022-1031(74)90070-5

Levon, E. (2016). Gender, interaction and intonational variation: The discourse functions of High Rising Terminals in London. Journal of Sociolinguistics, 20(2), 133–163. https://doi.org/10.1111/josl.12182

Nilsson, J. (2015). Dialect accommodation in interaction: Explaining dialect change and stability. Language & Communication, 41, 6–16. https://doi.org/10.1016/j.langcom.2014.10.008

Pichler, H. (2016). Introduction: discourse-pragmatic variation and change. In H. Pichler (Ed.), Discourse-Pragmatic Variation and Change in English: New Methods and Insights (pp. 1–18). Cambridge: Cambridge University Press. https://doi.org/DOI: 10.1017/CBO9781107295476.001

Raymond, C. W. (2018). On the relevance of dialect: Conversation analysis and contact linguistics. Journal of Sociolinguistics, 22(2), 161–189. https://doi.org/10.1017/CBO9781107415324.004

Yngve, V. H. (1970). On getting a word in edgewise. In Chicago Linguistics Society, 6th Meeting (pp. 567–578).

Gendered word-search strategies in Glasgow


Some non-lexical features of speech are known to vary according to gender (e.g. voice quality—Becker et al, 2014). While clicks are phonemically rare (Ladefoged, 1996), they are common as discourse markers in many languages, including English (see Ogden, 2013 etc.), and there is some evidence that they might vary according to social factors, but few studies investigate this to date (Pillion, 2018). This study proposes a method for handling interactional variables in gendered language through a variationist lens.

Previous studies indicate clicks in English can occur alongside an inbreath and creaky and oral or nasal particles. Clicks convey emotion (e.g. disapproval, disagreement, sympathy) or help manage sequences in interaction. Sequence-managing clicks can mark word search (e.g. the act of searching for a word in interaction), index the start of a new sequence, mark incipient speakership (the shift of one speaker to another), backchannel, and more (Wright, 2011; Ogden, 2013).  While some functions (e.g. indexing a new sequence) have been well-researched, others, such as clicks’ role in word search, remain unclear.

The only social factors that have been examined in conjunction with click production are region and style and only on a small scale (see Moreno, 2016). However, in Ogden’s study, the three speakers who produced the most clicks were female (2013), suggesting an interaction between click production and gender. This paper therefore aims to answer the following research questions:

  1. What is the phonetic form of word search clicks?
  2. Do male and female speakers construct word searches with clicks differently?

The study is based on over 20 hours of audio and visual recordings of same-gendered, self-selected dyadic interactions between Glaswegians ages 18 to 60. The sample draws from a larger study on click production and function, currently analysing around 400 tokens. 

A combination of interactional and quantitative methods were used due to the nature of studying interactional variables on a large scale. All word search sequences were identified and narrowly phonetically transcribed in Praat. Word searches were coded (1) for the presence of particles (e.g. er, um) and (2) phonetic accompaniments indicated by previous studies (e.g. inbreath, creakiness, nasality). Where clicks occurred, place of articulation was coded (bilabial to lateral, Ogden 2013). 

Preliminary results suggest a difference in the way male and female speakers construct word search sequences rather than a gender difference in the use of clicks. This places clicks in word search sequences in the wider context of gender-constrained linguistic variables and demonstrates the importance of examining interactional variables in context.



Becker, K., Khan, S. U. D., & Zimman, L. (2014). Voice quality variation across gender identities. The Journal of the Acoustical Society of America, 135(4), 2424 https://doi.org/10.1121/1.4878058

Ladefoged, P., & Maddieson, I. (1996). The Sounds of the World’s Languages. Oxford, UK: Blackwell.

Moreno, J. (2016). Tut-tut: A sociophonetic study of clicks in female speakers from three regions of Scotland. University of Glasgow.

Ogden, R. (2013). Clicks and percussives in English conversation. Journal of the International Phonetic Association, 43(3), 299–320. https://doi.org/10.1017/S0025100313000224

Pillion, Betsy, ‘English Clicks: Individual Variation in Speech Preparation and Stance Display’ (New York, USA, 2018)

Wright, M. (2011). On clicks in English talk-in-interaction. Journal of the International Phonetic Association, 41(July 2011), 207–229. https://doi.org/10.1017/S0025100311000144

Language and gesture in narrating migration: proposal for a multimodal corpus
PRESENTER: Chiara Meluzzi


This works presents the first results of a project devoted to safeguard the linguistic and cultural heritage of the Alpine area of Biella (Italy). This research has been carried out in a multidisciplinary but also multimodal way, through a linguistic-anthropological fieldwork devoted to language documentation of the migratory experiences by subjects coming from several regions of Italy after WWII until today. Thus, this sociolinguistic setting  provides an optimal situation of language contact between different Italian regional varieties and Italian dialects. The project aims at investigating both the preservation of original dialects and the acquisition of the new dialectal features typical of the Biellese dialect. Moreover, a special interested has been devoted to multimodality in narrating the migratory experience.

Data were collected through semi-structured interviews, recorded with both audio and video support. For the purpose of this presentation, we selected 9 speakers (4 female, 5 male) aged between 50 and 80 y.o., and corresponding to first generation immigrants in Biella from different Italian regions (namely, Veneto, Sardinia, Sicily, and Campania). The audio-visual recordings have been transcribed and annotated for POS and gestures through ELAN; for gestures we focused on gaze and eyebrows following the FACS protocol (Ekman 1982). Separately, after word tokenization in ELAN, we use PRAAT for a fine-grained phonetic transcription at both segmental level, and supra-segmental level by using the Momel-Intsint plugin (Hirst 2007).

First results show that some typical features of Sicilian Italian regional variety have been preserved, in particular for what it concerns the realization of the intervocalic alveolar fricative as voiceless [s], whereas Biellese, like many Northern varieties, show the typical voiced realization (Canepari 1979). Conversely, mid vowels seems to be subject to change in stressed syllables, perhaps in relation to the topic of the interview. Moreover, some typically northern realizations are identifiable, for example in the pronunciation of dental affricates some speakers tend to produce more voiceless affricates [ts] in the post-sonorant context,  where the southern variant would provide the voiced [dz].

As for gestures, a first analysis indicated that Southern speakers tend to gesticulate with their hands more than Northern speakers. However, in all speakers raising of eyebrows indicated the beginning of narration of a painful experience, in particular concerning migration. The direction of the gaze also changes during these narrations, moving away from the camera into a distant point behind the interviewer.

These preliminary results confirm the importance of multimodal analysis, especially in narration of immigrant experience. This research will also help raising awareness of heritage varieties and their preservation even within the same national border.



Canepari, Luciano (1979) Italiano standard e pronunce regionali, Padova: Cleup.

Ekman, P. (1982) Methods for measuring facial action, in K.R. Scherer & P. Ekman (eds.) Handbook of methods in nonverbal behavior research, New York: CUP, 45-135.

Hirst, D. J. (2007) A Praat plugin for Momel and INTSINT with improved algorithms for modelling and coding intonation. Proceedings of the XVIth International Conference of Phonetic Sciences, 1223-1236.

Articulatory variation in Icelandic: voiced fricatives and approximants accross registers


While Icelandic is thought to be a conservative language, cited pronunciations often do not reflect actual speech. In connected speech, the articulation of voiced fricatives and approximants (traditionally symbolized [v], [ð], [j] and [ɣ]) is varied. The sounds frequently weaken or fully delete (Árnason 2011, Helgason 1993). For example, even in somewhat clear and formal speech, [v] and [ɣ] weaken to [w] or delete following a rounded back vowel, e.g. mágur ‘brother-in-law’ and máfur ‘seagull’ are pronounced [mauwʏr̥] or [mau:ʏr̥] (Árnason 2011). Unsurprisingly, weakening and deletion is more frequent in informal or colloquial speech, where it can result in further deletion and the emergence of informal variants e.g maður ‘man’ [ma:.ðʏr̥] à[ma:.ur̥] à [ma:r̥] and náttúrulega ‘naturally’ [nauhturʏlɛɣa] à [nauhturʏlɛa]à [nauhtl̥a]. Deletion happens within a prosodic word, e.g. þú veist ʽyou knowʼ [θu.veist] à [θu.eist]/[θust]. Previous phonological research has mainly focused on areal and generational dialects (see Árnason 2011) and not connected speech processes (though see Helgason 1993) nor variation between formal and colloquial speech.

This paper reports the results of a study which looked at the articulation of voiced fricatives and approximants in formal and colloquial spontaneous speech. Data on formal speech comes from the Parliament Speech Corpus, which contains recordings of unprepared and unwritten speeches from discussion periods at the Icelandic Parliament in 2004-2005. This data should therefore reflect natural speech under formal conditions. To get an actual representation of natural colloquial speech, I created a corpus using data from an Icelandic podcast with two native speakers, a male and a female. Podcasts are relatively recent in Iceland. This kind of data has never been used before for Icelandic. The advantage is that this is a fully casual and informal conversation between speakers that are not trying to realize expected pronunciations in accordance to prescriptive rules. This is apparent e.g. from their usage of English slang and high frequency of discourse fillers and hedging words.

The results show that weakening and deletion happen both in formal and colloquial speech but is more extreme in the latter. It is most frequent intervocalically when following an unstressed syllable. Sound environment, however, is not the only significant factor. Another is frequency, e.g. [ɣ], in the adverbial suffix -lega [lεɣa] ‘-ly’, is weakened or fully deleted in formal speech. While in colloquial speech the suffix is consistently pronounced [la]. Furthermore, deletion leads to further reduction, e.g.  [nauhtl̥a] for náttúrulega ‘naturally’ (as in Helgason 1993). Grammatical function is also a significant factor, e.g. [θust] or [θst] for þú veist ‘you know’ in colloquial speech only occurs when the phrase is used as a discourse marker or hedger. These informal pronunciations have resulted in new written variants emerging in informal writing, e.g. on Twitter, some examples are náttla (from náttúrulega) and þúst (from þú veist).



Árnason, K. (2011) The Phonology of Icelandic and Faroese. Oxford: Oxford University Press.

Helgason, P. (1993) On Coarticulation and Connected Speech Processes in Icelandic. Reykjavík: Málvísindastofnun Háskóla Íslands

10:45-12:25 Session 2C: Panel Coherence
Location: Madrid
Coherence: Outcome or cause of language change?


Sociolinguistic theory characterizes language varieties – such as dialects and speech styles – in terms of shared linguistic attributes.  It also recognizes variation and change as inherent properties of language. Conjointly, these two principles raise the problem of coherence: how can variable attributes form a coherent language variety?  How does any variety remain systematic in the presence of linguistic change? The postulates of ‘orderly heterogeneity’ (Weinreich et al. 1968) and shared norms are advanced to account for the regular patterns of internal, social and stylistic distribution of individual linguistic variables, but there is little empirical or theoretical work on whether entire systems of variables cohere or covary. Every language has multiple variable features; do these all show similar regular patterning? Sociolinguistic work on indexical fields and identity suggests that variables may differ substantially in their social significance, and that speakers make intentional and idiosyncratic choices about the use of variables. This implies that idiolects may differ unpredictably and that community usage may not be very coherent beyond the level of single variables.

More nuanced questions arise with respect to the modular nature of linguistic structure. Is there greater coherence, i.e., similar patterning or covariation, among variables within a grammatical module or across modules? Is there a tendency for variability within a subsystem to maintain an internal equilibrium (e.g., balanced segment inventories, paradigmatic regularity)?  Do variables that seem to be ‘below the level of conscious awareness’ participate in the same kind of regular patterning as variables that are subject to explicit social evaluation?

Language change introduces another dimension to the question of coherence. Change typically proceeds through synchronic variation, but the social and structural significance of the new variants will likely not cohere with existing variables. On the structural side, innovations may trigger other changes within a (sub)system, which can have the effect of restoring coherence, e.g., the first change in a chain shift. Externally triggered changes, arising from language or dialect contact, can potentially introduce structural variation that is substantially different from existing variables, e.g., phonemic neutralizations, novel word orders. Socially, externally triggered changes will typically have different social distributions and indexicalities from established variables, and hence would be unlikely to covary with them, disrupting the coherence of the variety.

To further explore notions of coherence, the panelists in this workshop will address such questions as:

  • How do we define coherence? How does it fit into a holistic theory of linguistics and language variation and change?
  • To what extent and in what areas of the grammar is systematic covariation and coherence evident? Do variables show markedly different social indexicalities and distributions?
  • How does language change disrupt or converge with existing patterns of coherence among variables in variety?
  • Do contact varieties exhibit less coherence, especially if there is a constant influx of new speakers and new features?
  • How is coherence reflected in language acquisition and across the life-span? Are speakers more or less coherent in childhood and adolescence? Does coherence increase or decrease later in life? 
Not anything goes: On coherence in language and the penalty for being incoherent


The research reported in this paper is part of a larger endeavour to model the structure which underlies variability in languages. The beginning of the 21st century witnesses a remarkable functional elaboration of intermediate varieties between dialect and standard language, which are often regarded  as a threat to the uniformity of both dialect and standard language. In this paper, we claim that there is much more uniformity in present-day (standard) language than generally assumed. We report data from a production and a perception study in Flanders (Belgium).

A corpus analysis of the speech of 10 West-Flemish females in three communication situations revealed compelling evidence for rigid patterning in the colloquial variety dubbed “Tussentaal”: rather than a random, idiolectal mix of dialect and standard features, this intermediate variety turned out to be structured by identical “implicational scales” across speakers: the presence of specific dialect features was found to be a reliable predictor for the presence of other features, whereas the presence of the latter features did not automatically imply the presence of the first (Ghyselen & Van Keymeulen 2016).

In a follow-up perception experiment, this implicational coherence turned out to be more than a production fact. 44 female adolescent participants rated 6 experimental samples featuring West-Flemish dialect or near-standard Dutch, which were either implicationally coherent (4 samples) or incoherent (2 samples); samples were evaluated both on scales pertaining to (amongst others) artificiality, and on an open response item. On the latter, the incoherent dialect fragments in particular were downgraded as “fake”, “unnatural”, “artificial”, “incompetent”… by nearly half of the participants. (In)coherence, as a consequence, is a production reality which is strongly entrenched in hearer perceptions. 

Time permitting, we will account for the noticeable downgrading of implicationally incoherent language in terms of disharmony, viz. the fact that the “wrong” variants in the incoherent samples do not index social meanings which are beneficial to a “harmonisation” of the speaker’s and the hearers’ identities  (Grondelaers & Van Hout 2015).



Ghyselen, A-S. & J. Van Keymeulen (2016). Implicational scales in colloquial Belgian Dutch. Dialectologia et Geolinguistica. Journal of the International Society for Dialectology and Geolinguistics 24: 62-82.

Grondelaers, S. & R. van Hout (2015). How (in)coherent can standard languages be? A perceptual perspective on co-variation. Lingua 172-173 (Theme issue Coherence, covariation and bricolage. Various approaches to the systematicity of language variation, ed. by Frans Hinskens and Gregory R. Guy), 62-71.

Modelling lectal coherence: The case of Swabian German


Forty years ago, Weinreich, Labov, and Herzog (1968:188) observed that “idiolects do not provide the basis for self-contained or internally consistent grammars”, rather it is the grammar of the speech community, governed by social factors, which reflects regularity and coherence and where linguistic change occurs. Guy and Hinskens (2016:1-2) claim that the concept of orderly heterogeneity implies that “speech communities are sociolinguistically coherent .... [meaning that] the community should collectively behave in parallel: variants (or rates of use of variants) that index a given style, status, or a social characteristic should co-occur ... like a falling domino can make a row of neighbouring dominoes fall” and “can be characterized as displaying coherence.”

Co-variation is one approach to coherence; however, another and perhaps more promising approach utilises Guttman (1944) and Bickerton (1973) implicational scaling to identify types and levels of coherence across lectal chains in an implicational-like pattern. In their study of the Belgian dialect Tussentaal, Ghyselen and Van Keymeulen (2016:15) argue that Tussentaal “is not just a random idiolectal mix of dialect features, but that it is structured by implicational principles shared across the speech community.”

To explore the concept of sociolinguistic coherence and how it shapes variation and fosters or constrains language change, this paper examines language usage in two Swabian speech communities across two points in time (1982 and 2017). Following the traditional quantitative variationist approach, pioneered by Labov (1966) in analysing the variation between dialect and standard language features, coupled with implicational scaling, and drawing on concepts from the order and lattice theory of mathematics, ten phonological and morpho-syntactic features of Swabian, along with five social factors (speaker age, sex, education, orientation, mobility) are modelled to examine aspects of linguistic coherence and language change across the two time periods. The hypothesis of this research is that more coherent lects are less vulnerable to change and convergence to the standard language, while less coherent lects are more susceptible. The modelling approach used brings together three views of lectal coherence – covariation, implicational scaling, and lattice theory – to demonstrate a holistic approach to the theory of linguistic coherence and its impact on language change.



Bickerton, Derek. (1973). The nature of a creole continuum. Language, 49(3), 640–669.

Ghyselen, Anne-Sophie & Jacques Van Keymeulen. (2016). Implicational scales in colloquial Belgian Dutch. Dialectologia et Geolinguistica. Journal of the International Society for Dialectology and Geolinguistics 24:62-82.

Guttman, Louis. (1944). A Basis for Scaling Qualitative Data. American Sociological Review, 9(2), 139–150.

Guy, Gregory R., & Hinskens, Frans. (2016). Linguistic coherence: Systems, repertoires and speech communities. Lingua, 172–173, 1–9. http://doi.org/10.1016/j.lingua.2016.01.001

Labov, William. (1966). The Social Stratification of English in New York City. The Social Stratification of English in New York City (2nd ed.). Cambridge: Cambridge University Press.

Weinreich, Uriel, Labov, William, & Herzog, Marvin I. (1968). Empirical foundations for a theory of language change. In W. Lehmann & Y. Malkiel (Eds.), Directions for Historical Linguistics. Austin: University of Texas. (pp. 95–188).

Language change in real-time: Life span coherence in individual repertoires. Results from a panel study in Ulrichsberg / Austria


Individual repertoires are always more or less focussed (Le Page & Tabouret-Keller 1985). Due to the process of “Entdiglossierung” (Bellmann 1983) it is generally assumed, that Bavarian language reality is diaglossic and repertoires therefore rather diffuse (Auer 2001). Yet there is counter-evidence for more focussed varieties in this area (e.g. Ender & Kaiser 2014; Vergeiner 2018)—it has been shown that the use of dialect- respectively standard-features (at least for some speakers) are clearly constrained both functionally and structurally (viz. bipolar co-occurrence restrictions and diglossic covariation patterns exist). After all, more in-depth studies on this issue are required—especially on the question, whether there are inter-individual and group differences in the repertoire-structures within one region and additionally, to what extent one has to take into account intra-individual developements—for example between early and late adulthood.

This paper investigates these questions for rural Austria and the village of Ulrichsberg (Upper Austria). Based on a real-time panel study with data points in 1977 and 2017 (see Scheutz (1985) for the 1977 data; DiÖ and Wallner (in prep.) for the 2017 survey) it seeks for differences in the repertoires of 14 speakers, which represent diverging social classes (farmers, lower and upper middle class). The study is based on two types of conversational data (formal interviews / informal talks), which are quantitatively analysed. Thereby mainly phonological variables are focused (derounding, word final /ç/-deletion, etc.).

The analysis seeks on the one hand for differences in the distribution of the variables in the different settings and groups (viz. diastratic and diaphasic covariation of the variants), but on the other hand also for co-occurrence relations for selected variants and speakers by conducting a co-occurrence-analysis (for this quantitative method see e.g. Vergeiner 2018). By that, the functional and structural separation of the different variants—and therefore varieties—is spotted. Comparing the results for 1977 und 2017, it will be detected whether the coherence in the repertoires remained stable or change took place.



Auer, Peter. 2001. Zur soziodialektologischen Bewertung des Standard-Dialekt-Ausgleichs. In Kurt Egger & Franz Lanthaler, Die deutsche Sprache in Südtirol: Einheitssprache und regionale Vielfalt, 27–39. Wien: Folio.

Bellmann, Günter. 1983. Probleme des Substandards im Deutschen. In Klaus Mattheier, Aspekte der Dialekttheorie, 105–130. Berlin: de Gruyter.

Deutsch in Österreich (=DiÖ) / German in Austria. 2017. https://dioe.at/en/

Ender, Andrea & Irmtraud Kaiser. 2014. Diglossie oder Dialekt-Standard-Kontinuum: Zwischen kollektiver, individueller, wahrgenommener und tatsächlicher Sprachvariation in Vorarlberg und im bairischsprachigen Österreich. In Claudia Bucheli Berger & Dominique Huck, Alemannische Dialektologie: Dialekte im Kontakt, 131–146. Stuttgart: Steiner.

Le Page, Robert B. & Andrée Tabouret-Keller. 1985. Acts of identity: Creole-based approaches to language and ethnicity. Cambridge: CUP.

Hannes Scheutz. 1985. Strukturen der Lautveränderung. Variationslinguistische Studien zur Theorie und Empirie sprachlicher Wandlungsprozesse. Wien: Böhlau.

Vergeiner, Philip C. 2018: Kookkurrenz - Kovariation - Kontrast: Formen und Funktionen individueller Dialekt‑/Standardvariation in Beratungsgesprächen an der Universität Salzburg. Wien: Lang.

Wallner, Dominik. in prep. Sprachwandel in Realtime am Beispiel von Ulrichsberg. Dissertation: University of Salzburg.

10:45-12:25 Session 2D: Panel Dialectometry
Location: Moscow
Dialect classifications in Europe revisited


1. A general overview

Most of the dialect classifications of European languages have been made in the framework of neogrammarian's and using isoglosses as the unique tool for drawing dialect boundaries. Nevertheless, Linguistics in general and the study of linguistic variation in particular have changed enormously in the last century. Nowadays linguistic researches are developed in different frameworks, apart from the diachronic, in the synchronic point of view there are diverse frameworks such as structuralism (Weinreich 1954) and different generative approaches (Barbiers 2010). It is assumed that dialect constitutes a linguistic system; so, when dialects are compared, different linguistic systems are compare, and non linguistic features of each dialect. It seems that, on the one hand, subjectively chosen single features based dialect classifications must be cast aside, and it is considered as a better choice to use great amount of data or systemic features of dialects. On the other hand, it is optional or compulsory to take into account the dialect as a linguistic structure, and does it mean that whether it is not a system change between varieties, there is not new dialect? This approach poses more than one question: Is it compulsory to analyze linguistically the data? And a more problematic question: When could we consider that there has been linguistic system change?

Meantime, the study of linguistic variation has developed: on the one hand, the knowledge of the areal variation has been improved hugely with the apparition of linguistic atlases in most of the languages of Europa; on the other hand, the develop of the sociolinguistic variation allows the analysis of the extralinguistic factors on the linguistic variation. Moreover, nowadays is possible to go further due to the implementation of statistic analysis for geolinguistic variation (cluster analysis, multidimensional Scaling, etc.). With this technological improvement all language variation, areal or sociolinguistic, can be analyzed by figures and maps, quantifying all linguistic differences and linguistic variation.

Despite the fact that either the knowledge of the data has grown exponentially or the massive use of technological tools for determining linguistic areas and social class are better known, there are still some questions to be resolved. One of them will be that neither areal nor social dialectology can determine where finishes a linguist system and where does a new one begin (Kretzschmar 2000:236). Another question will be which name would be given to those linguistic areas. Among the pending issues, and without being exhaustive, the distinction between boundary and dialect continuum is still unresolved, for example. As far as the boundaries of dialect is concerned innovative suggestions of the topic of the gradualness of the linguistic differences among dialect areas is crucial.

It seems that it is a good moment to review the dialectal classifications that have been made up to now in Europe. It is important to know which methodology has been used to classify dialects, the framework of the classifications, the advantages and disadvantages of each classification, the handicaps of them, etc. To sum up, it looks like it is the moment to establish the state of the art of the dialectal classifications in the European languages.

The panel is closely related to the conference's theme, as it proposes the review of dialect classifications in all languages of Europe, including minority languages (Basque, Galician, Catalan, etc.).

The rationale to propose this panel is to promote the discussion about the methods that have been used until now and are used even nowadays to classify dialects in Europe.



Weinreich, U. (1954): Is a Structural Dialectology Possible? Word 10:2-3, 388-400, DOI: 10.1080/00437956.1954.11659535

Barbiers, S. (2010). Language and space: Structuralist and generative approaches. In P. Auer & J. E. Schmidt (eds.): Language and Space An International Handbook of Linguistic Variation. Volume 1: Theories and Methods: 125-142. Berlin/New York: De Gruyter Mouton.

Kretzschmar, W. A. Jr. (2000): The Future of Dialectology. Leeds Studies in English 30: 271-288.



2. Aims of the panel:

  • To contrast the methods used to classify dialects in distinct European languages.
  • To discuss about distinct dialect boundaries
  • To discuss about the ways to organize dialect boundaries hierarchically
  • To discuss about the features of distinct dialect boundaries
  • To discuss about the possibility to work together in a future project about the requirements of            dialect classification


3. Key discussion questions

  • Dialect boundaries
  • Ways to organise dialect boundaries hierarchically
  • Features of distinct dialect boundaries
  • Possibility to work together in a future project about the requirements of dialect classification


4. Organization of the panel

The panel will have 9 slots, and will be developped in one day, as follows:

  • 9:00-9:30: introduction to the panel (Aurrekoetxea & Ensunza)
  • 9:30-11:00: contributions
  • 9:30-9:50: G. Brun-Trigaud (U. Côte d’Azur, CNRS, BCL), Clément Chagnaud (U. Grenoble Alpes, CNRS, Grenoble INP, LIG) & Philippe Garat ( U. Grenoble Alpes, CNRS, Grenoble INP, LJK): Identification of lexical areas templates throughout the Occitan domain
  • 9:50-10:10: X. Sousa & F. Dubert  (ILG-US Compostela): Internal and external language borders in the Galician domain
  • 10:10:10:30: L. Unamuno (UPV/EHU): Classifications of Basque dialects
  • 10:30-11:00: discussion


  • 11:00-11:30: coffee break
  • 11:30-13:00: contributions
  • 11:30-11:50: E. Carrilho (Univ. Lisbon): On the classification of Portuguese dialects
  • 11:50-12:10: R. d'Andrés (Univ. of Oviedo): Challenges in the taxonomy of linguistic diversity and  dialectal classification. The Asturian case    
  • 12:10-12:30: P. Roseano (Univ. de Barcelona, U. of South Africa) & F. Finco (Pädagogische Hochschule Kärnten): Classification of Friulian dialects
  • 12:30-13:00: discussion


  • 13:00-14:30: lunch


  • 14:30-16:00: contributions

- 14:30-14:50: D. Mikulėnienė (The Institute of the Lithuanian Language): Development of Lithuanian dialectal classification across centuries

- 14:50-15:10:  Jožica Škofic(ZRC SAZU, Fran Ramovš Institute of the Slovenian Language): Dialect classification of Slovene-history and perspectives

- 15:10-15:30: M-P. Perea (U. of  Barcelona): Catalan dialect classification revisited.

- 15:30-16:00: discussion


  • 16:00-16:30 coffee break
  • 16:30-17:30: contribution
  • 16:30-16:50: A. Stafecka (Latvian Language Institute of the University of Latvia): Latvian dialects and their classification
  • 16:50-17:10: A. P. Versloot (University of Amsterdam/Fryske Akademy): Review of the dialectal classifications in Europe
  • 17:10: 17:30: discussion


  • 17:30-18:30: final discussion and proposals for the future.

So, there will be between 11 papers. As far as is concerned to the discussion format two types of discussion are foreseen: The first discussion type will be discussion about concrete papers, after each set of contributions (4 discussions). The second discussion type will be carried out at the end of the panel and it will deal with the panel topic and future steps.

We have some other invited contributors that have agreed to participate, but we still lack final confirmation. If the panel will be accepted, we think that it will be interesting to make a special effort to include more languages: Frisian (we are in contact), Gaelic, Finnish...


5. Invited contributors

Maria Pilar Perea, University of Barcelona: mpilar.perea@ub.edu

Xulio Sousa, ILG-US Compostela: xulio.sousa@usc.es

Francisco Dubert, ILG-US Compostela: francisco.dubert@usc.es

Ramón d'Andrés, University of Oviedo: randrsd@uniovi.es

Guylaine Brun-Trigaud, U. Côte d’Azur, CNRS, BCL: Guylaine.Brun-Trigaud@unice.fr

Clément Chagnaud, U. Grenoble Alpes, CNRS, Grenoble INP, LIG

Philippe Garat, U. Grenoble Alpes, CNRS, Grenoble INP, LJK

Ernestina Carrilho, University of Lisbon: ernestina.carrilho@campus.ul.pt

Paolo Roseano, U. de Barcelona, U. of South Africa: paolo.roseano@ub.edu

Fanco Finco, Pädagogische Hochschule Kärnten: franco.finco@ph-kaernten.ac.at

Danguolė Mikulėnienė, Institute of the Lithuanian Language: dangoule.mikuleniene@gmail.com

Arjen P. Versloot, University of Amsterdam/Fryske Akademy: a.p.versloot@uva.nl

Lorea Unamuno, University of Basque Country (UPV/EHU): lorea.unamuno@ehu.eus

Jožica Škofic, ZRC SAZU, Fran Ramovš Institute of the Slovenian Language: guzej@zrc-sazu.si

Anna Stafecka, Latvian Language Institute of the University of Latvia

Catalan Classification Revisited


According to the latest dialectal division proposal (cf. Veny 1984), the Catalan language is divided into six main dialects: Rosellonés, Eastern Catalan, Western Catalan, Valencian, Balearic and Alguerés. The first five each contain subdialects which present more closely cohesive dialectal features.

The Valencian dialect has three subdialects called, by geographical location from north to south: Northern Valencian, Central Valencian (or “apitxat”) and Southern Valencian. The Balearic subdialects are associated with their particular islands: the varieties of Majorca, Minorca and Ibiza. The last one is joined to the variety of Formentera, not always with the approval of the speakers of that small island. The Eastern Catalan dialect, called in fac, “Central Catalan”, is divided into five subdialects: the Barcelona and Tarragona subdialects, the transitional Northern subdialect, and, finally, the so-called “xipella” and “salat” subdialects. In this case, unlike the geographically defined Valencian and Balearic subdialects, the “Central” Catalan subdialects are defined by different dialectal characteristics. Some of them follow a demographic approach (the Barcelona and Tarragona subdialects), with the former especially having an integrative and an irradiator character. The transitional Northern subdialect marks a transitional area between the Eastern Catalan at the South and Rossellonés in the North. Finally, “xipella” and “salat” subdialects are characterized by only a single dialectal feature.

The first proposal of Catalan division (Milà i Fontanals 1861) was based on a phonetic distinction and since then different criteria have been used to divide the Catalan speaking area into several dialects and subdialects. Thus, the geographical situation, unique phonetic or morphological features, and some sociolinguistic traits have been mixed together to obtain a classification that seems unquestioned and unquestionable. It has two purposes: a) to be justified by using historicist criteria; and b) to maximize the linguistic unity of the Catalan language.

The aim of this paper is to explore the origin and history of the different proposals for Catalan dialect classification and to question not only the resulting number of dialects and subdialects but also their names. It is argued also that the use of Central Catalan is a marked denomination that gives preponderance to this dialect, which is closely related to the standard Catalan language. This name is also opposed to the other linguistic Catalan varieties, which, because of their distance from the prescriptive grammar, exhibit a variable degree of stigmatization. Together with the reviewing of the proposals, the validity of a number of criteria for dialect classification will be examined.



Milà i Fontanals, M. (1861) De los trovadores en España. Barcelona: Librería de Joaquín Verdaguer, 462.

Veny, J. (1984) Els parlars catalans. Palma de Mallorca: Moll

Classification of Friulian dialects


The beginnings of the study of dialectal variation of the Friulian language can be traced back to Ascoli’s (1873) Saggi Ladini. However, it is only in the twentieth century that this discipline experienced a greater development, which resulted in the publication of a considerable number of linguistic atlases including data for Friulian, but very few attempts to classify the dialects of that language.

The most important classifications were put forward by Francescato (1966) and Frau (1984), who present a substantially concordant description of the dialects of Friulian. Other works by the same two authors (Frau 1989; Francescato 1965) and others (Iliescu 1972; Marcato 2001; Vicario 2005; Heinemann 2007; Roseano 2015, i.a) agree with such classification.

Both Francescato (1966) and Frau (1984) base their classifications of Friulian dialects on a limited number of isoglosses. Such isoglosses correspond to about fifty linguistic features to which the authors in question attribute particular importance. Such features have to do mainly with phonology and, to a lesser extent, with morphology and lexicon (while other aspects of the language, such as syntax, are not taken into consideration).

Another critical aspect in the studies about the classifications of Friulian dialects is the limited number of attempts to use dialectometric techniques (Lazard 1985; Goebl 1988; Bauer 2010, i.a.). The lack of dialectometric analysis is due to the fact that the most important linguistic atlas for Friulian, the Atlante storico-linguistico-etnografico friulano (Pellegrini 1972-1986), has not been transformed into a computer database yet.



Ascoli, G. I. (1873) Saggi ladini. Archivio Glottologico Italiano 1, 1-556.

Bauer, R. (2010). Verifica dialettrometrica della Ladinia di Graziadio Isaia Ascoli (a 100 anni dalla sua morte). In Maria Iliescu/Heidi Siller-Runggaldier/Paul Danler (ed.): XXVe CILPR Congrès International de Linguistique et de Philologie Romanes. Innsbruck, 3-8 septembre 2007: vol. 7: 3-10. Tübingen: Niemeyer.

Francescato, G. (1965). Uno studio sulla dialettologia del Friuli. In Albert Joris van Windekens (ed.): Communications et rapports du Premier Congrès International de dialectologie générale Louvain 1960, vol. 4: 122-129. Louvain, Centre international de dialectologie générale.

Francescato, G. (1966) Dialettologia friulana. Udine: Società Filologica Friulana.

Goebl, H. (1988): Il posto dialettometrico che spetta ai punti-AIS 338 (Adorgnano, Friuli), 398 (Dignano/Vodnjan, Istria) e 367 (Grado, Friuli). Presentazione di tre carte di similarità. Linguistica 28: 75-103.

Heinemann, S. (2007) Studi di linguistica friulana. Udine: Società Filologica Friulana.

Iliescu, M. (1972) Le frioulan à partir des dialectes parlés en Roumanie. Den Haag/Paris: Mouton.

Lazard, S. (1985): Les frontières du dialecte frioulan: étude dialectométrique à la lumière de la méthode globale dʼHenri Guiter. Revue de Linguistique Romane 49: 27-70.

Marcato, C. (2001) Friuli Venezia Giulia. Roma/Bari: Laterza.

Pellegrini, G. B. (ed.) (1972-1986) Atlante storico-linguistico-etnografico friulano, 6 vol. Padova: Istituto di glottologia dell’Università di Padova.

Roseano, P. (2015). Suddivisione dialettale del friulano. In Heinemann, S. & Melchior, L. (eds.): Manuale di linguistica friulana: 155-186. Berlin: De Gruyter Mouton.

Vicario, F. (2005) Lezioni di Linguistica Friulana. Udine: Forum.

Classifications of Basque dialects


In traditional dialectology the classification of dialects has been done using only a few features selected by dialectologists according to their criteria and unfortunately this type of classification has been accepted for a long period of time. However the publication of linguistic atlases and the appearence of quantitative techniques to classify dialects have created an appropiate situation to explore and propose new classifications.

As for the Basque dialects concern, the first feature-based scientific classification was carried out by Bonaparte (1869). He classified Basque dialects into 8 dialects (Biscayen, Guipuzcoan, Northern Upper Navarrase, Southern Upper Navarrase, Labourdin, Western Low Navarrase, Eastern Low Navarrase and Souletin), 25 subdialects and 50 varieties. Half a century later, Azkue´s (1905) classification was 6 dialects (Biscayen, Guipuzcoan, Upper Navarrase, Labourdin, Low Navarrase, Souletin and the Roncalese) and newly, Zuazo (1998, 2010) sets in 6 Basque dialects (Western, central, Navarrese, Eastern Navarrese, Navarrese-Labourdin and Souletin). All of them are important classifications in the traditional Basque dialectology, nevertheless it is not known what techniques have been used to classify nor what criteria to denominate dialects, subdialects and varieties.

In the 21st century, though, corpus-based classifications become reality. The first dialectometric application to the Basque was about the classification of Navarrese dialects (Aurrekoetxea, 1992). Afterwards, other corpus have been used; Bourciez (Aurrekoetxea & Videgain 2009), EAS-Corpus of the Geo and Sociolinguistic Atlas of Basque (Aurrekoetxea & Ormaetxea 2006,) and EHHA-Corpus of the Basque Linguistic Atlas (Aurrekoetxea et al. 2018). The novelty incorporated in these corpus-based classifications is, on the one hand, the handling of a large amount of data and, on the other, new techniques to analyze and to classify the data. Therefore, it is a matter of answering to what these new classifications are due to the fact that the dialects have changed or if the handling of a large amount of data and new techniques have an influence.



Aurrekoetxea, G. (1992): Nafarroako euskara: azterketa dialektometrikoa. Uztaro 5: 59-109.

Aurrekoetxea, G. & J. L. Ormaetxea (2006): Research project - Socio-geolinguistic atlas of the Basque language. Euskalingua 9: 157-163.

Aurrekoetxea, G., Ormaetxea, J.L., Videgain, X. (2018): Euskalkien sailkapen zientifikoa: lexikoa (1). In L. Unamuno, A. Romero, A. Etxebarria & A. Iglesias (eds.): Linguistic Variation in the Basque and Education-III / Euskararen bariazioa eta bariazioaren irakaskuntza-III: 126-140. Bilbo: UPV/EHU.

Aurrekoetxea, G. & Videgain, X. (2009): Le projet Bourciez: Traitement géolinguistique d’un corpus dialectal de 1895. Dialectologia 2: 81-111.

Azkue, R.M. de. (1905) Diccionario vasco-español-francés. Bilbao.

Bonaparte, L.L. (1869) Carte des sept provinces basques, montrant la delimitation actuelle de l’euscara. London.

Zuazo, K. (1998): Euskalkiak, gaur. Fontes Linguae Vasconum 78: 191-233.

Zuazo, K. (2010) El euskera y sus dialectos. Iruñea: Alberdania.

10:45-12:25 Session 2E: Panel Minority
Location: Paris
European minority and diaspora languages


This panel explores variation in European minority and diaspora languages. We consider a diaspora language to be one spoken by people who have resettled in an area outside of their linguistic community. These speakers may have diverse origins and could come from different communities, social strata and even nations. Comparatively, a minority language is a variety, or a cluster of varieties, that is historically spoken in a particular region where another language—itself potentially unique from an official majority language—is predominantly spoken.

There is overlap between diaspora and minority languages. Many of the languages in our panel can be described as both diaspora and minority languages (Yiddish, Wymysiörys, Pomeranian, Greko). Whereas others are only diaspora languages, but not, strictly speaking, minority languages (Spanish spoken amongst Amerindian communities, Italo-Romance and Dalmatian varieties spoken in the Americas); the remainder being minority but not diaspora languages (Sorbian, Nahuatl). These issues entail a gamut of political repercussions. Consider how the “European Charter for minority or regional languages” defines minority and regional languages as those languages traditionally used within a given territory of a state by nationals of that state who form a group numerically smaller than the rest of the state’s population and which are different from the official language(s) of that state, including neither dialects of the official language(s) nor migrant languages. The charter protect some diaspora languages, whereas others (including Yiddish, Ladino and Romani) are referred to as “non-territorial languages”; i.e., languages used by nationals of the state which differ from the language(s) used by the rest of the state’s population but which, although traditionally used within the state’s territory, cannot be identified with a particular area thereof.

Aside from the geographic array, the linguistic variation explored attests to phenomena at many levels. This includes phonetic/phonological variation, relating to historical sound changes in Yiddish and Sorbian morphological and morphosyntactic variation attested in Israeli and US American variants of Yiddish, contact phenomena influencing the expression of grammatical mood in Spanish varieties in contact with Quechua and Aymara, and an array of morphological phenomena in Italo-Romance and -Dalmatian varieties spoken in the Americas The contribution on Pomeranian is dedicated to developing an analysis of the syntactic structure of this language which accounts for language change. Further, the contributions also extend beyond specific grammatical phenomena and also touch upon anthropological issues relating to verbal art for Sorbian, sociolinguistic and ethnolinguistic identity (for Wymysiörys, and Greko/Griko, and phenomena relating to language contact between minority and majority languages in a new sociological setting at different grammatical levels, and whether it results in complexification (as for Wymysiörys in contact with varieties of German and Polish, structural reduction (as for Zeelandic-Flemish in contact with Brazilian Portuguese, or other kinds of change as in Spanish in contact with Nahuatl, Aymara, and Quechua, among the many other examples in these contributions.

Contact and linguistic change in American and Israeli ultra-Orthodox Yiddish


Yiddish (iso yid) is maintained today only in some ultra-Orthodox communities in the US, Israel and Europe. Yiddish is used inside these communities as a minority language, while the majority language (mainly English, Hebrew or Dutch) serves for communication with the outside world. These communities offer a unique opportunity to observe the impact of language contact on linguistic change, due to their shared linguistic origin and similar sociolinguistic settings. In this talk, I analyze American and Israeli Yiddish varieties that (1) derive from the same Eastern European Yiddish dialects, from which they broke off at the same time, after WWII, and (2) are spoken in small, segregated, highly traditional, and close-knit Hasidic communities. They differ, however, in location and therefore in contact settings, so that the main factor distinguishing them is the identity of the majority language (American English or Israeli Hebrew). The impact of the majority languages has led to salient phonological and lexical differences between contemporary American and Israeli varieties. In addition, these Yiddish varieties also differ morphologically and syntactically.

The study examines several morphological and syntactic differences between Yiddish varieties, based on analysis of a spoken corpus consisting of Yiddish lectures, speeches, lessons and radio interviews of Israeli and American Hasidic speakers, recorded between 2007 and 2015 (Assouline 2017, p. 27).  

The comparative analysis demonstrates that, whereas Israeli Yiddish maintains the basic syntactic features of Yiddish (e.g. verb-second word order), American Yiddish speakers’ constant replication of English use patterns led to syntactic changes (e.g. allowing v3). The gradual structural convergence of American Yiddish with English, versus the maintenance of (documented) Yiddish syntax in Israel, arguably demonstrates the role of cognate lexical and grammatical material in related languages in enabling structural convergence.



Assouline, Dalit. 2017. Contact and Ideology in a Multilingual Community: Yiddish and Hebrew among the Ultra-Orthodox. Berlin: de Gruyter Mouton.

Evaluating linguistic variation in conditions of sparse data: Sorbian


The severely endangered minority Sorbian languages (iso hsb, dsb), endemic to the Eastern part of Germany, are dramatically under-researched. This lack of research extends from basic knowledge about speaker numbers, competence, and transmission to include also core aspects of linguistics, like phonology, morphology and syntax.

Having experienced centuries of marginalization, Sorbian texts are (sparsely) attested only from the 16th century. This makes evaluation of variation especially difficult, since the variation might reflect the state of the language, for example, a special dialect (our default assumption), but it might also be caused by other factors such as oral traditions and folksongs (which unfortunately have not been preserved in their original state either and are therefore hard to evaluate).

In this talk, for the first time we compare old Sorbian texts to folksongs[1], applying knowledge about neighbouring Germanic and Celtic literatures. From the linguistic side, results lend greater insights into historical sound changes in Sorbian, from the cultural side, we learn about aesthetic concerns of verbal art in this language, which, in turn shed light on a range of linguistic phenomena beyond sound patterns.

[1] We rely mainly on Smolers collection, which consists over 500 songs mainly collected by himself and Leopold Haupt, published as a book in Grimma 1841/43. It is available in (re)printed form and has never been analysed extensively.

Aymara and Quechua grammatical mood in Castellano Andino


Aymara (iso aym) and Quechua (iso que) are agglutinative languages spoken in multilingual Andean communities. These two languages are not only in regular contact with each other, but also with local varieties of Spanish, known collectively as Castellano Andino (hereafter: CA).

As little work is dedicated to how minority languages impact majority ones, we examine how Aymara and Quechua have impacted CA mood, a term we use to refer to grammatical(ized) modality in general. Mood is interesting because mood-marking and modal expressions play a key role in Andean discourse. Regional varieties of CA reflect Aymara and Quechua mood, even in the speech those who do not speak either indigenous language.

We show how CA reflects Aymara and Quechua mood by illustrating the emerging strategies used to express the modal values expressed in Aymara and Quechua grammar on different structural levels, including:

  1. tense/evidential paradigms: The evidential distinction of personal-knowledge vs. non-personal knowledge, for example, is based on the reinterpretation of an existing grammatical opposition: the CA perfect and pluperfect tense. Thus, the inflectional marking is maintained but the function is changing towards a mood rather than a tense distinction. This contact-induced change is hardly noticeable on the structural surface.
  2. incipient grammaticalization of new grammatical markers in CA, the evidential reportative dice/dice que/dizque: There is no direct transfer of a grammatical element from Quechua/Aymara into Spanish, but a contact induced step-by-step development of a new marker, based on the potential of the CA construction to express reported information in the appropriate contexts. There is also evidence of the reverse influence of CA dice, dice que on some varieties of Aymara and Quechua, where new reportative strategies emerge modeled on CA.
  3. lexical (and discoursive) strategies to express Quechua/Aymara mood marking, mainly by adverbs: as for seguramente to express inference/conjecture.
  4. boomerang development of conditionals: CA progressives influenced Quechua/Aymara constructions and now these constructions again influence CA.
  5. future tense uses are clearly separated: the analytic future is used as a tense, the Spanish synthetic future is modal (asking for permission/confirmation)

We show that contact induced change arises from multiple impulses, including:

  • existing potentialites of the grammatical structures and constructions in CA
  • the need to express and codify concepts of significant importance for the cultural and communicative routines of the speakers
  • lesser standardization pressure



Dankel, P. & Soto Rodríguez, M. (2012). Convergencias en el área andina. Neue Romania 41 (El español de los Andes), 89-120.

Hardman-de-Bautista, M. J. (1982). The mutual influence of Spanish and the Andean languages. Word, 33(1-2), 143-157.

Lee, T. Y. (1997) Morfosintaxis amerindias en el español americano. Desde la perspectiva del quechua. Madrid: Ediciones del Orto, Universidad Complutense de Madrid.

10:45-12:25 Session 2F: Panel Social Media
Location: New York 3
SoMe in Dialect Heaven: Norms and social meanings of Norwegian dialects online


Recent studies suggest that the use of written dialect features online is quite widespread in Norway (Evjen 2011; Rotevatn 2014; Røyneland 2018). However, the use of dialect seems to vary substantially between different age groups and geographical areas. A Gallup poll initiated by the Norwegian Language Council found that younger people report to use dialect to a greater extent than older users. Furthermore, people in the eastern part of the country report significantly less dialect use than people in other regions. Whereas only 16 % of the Oslo respondents report dialect use, as many as 45 % in Northern Norway report the same. This may have to do with the relative structural and/or perceived distance between dialect and the written standards (Bokmål and Nynorsk).

Dialects are relatively more prestigious – and less stigmatized – in Norway than in many other European countries, in fact to the extent that Norway has been described as a sociolinguistic paradise (Trudgill 2002). Dialects are spoken by people in all layers of society and within all social domains. Given that interaction on many CMC platforms share several of the characteristics of oral communication, it is to be expected that writing online gravitates towards the shape of spoken language, and for many Norwegians this simply means dialect.

The interesting question, then, is how dialect is used on different platforms and for what purposes. In our talk we address both questions by looking at data collected from three different platforms that differ with regard to their degree of publicness/privacy. We investigate what writing dialect actually means, and how norms for dialect writing are negotiated online, and also whether we find any correlation between level of formality and perceived audiences. We examine which features are used, how they are used, and what indexical meanings they can take on in different contexts (Johnstone & Kiesling 2008). We also analyse metalinguistic comments about dialect use, and hence, expressed implicit or explicit attitudes and ideologies. Our study shows that dialect features may be used to voice quite distinct, even opposing, attitudes and convictions, and that dialect use may get heavily policed, hence challenging the idea of Norway as a dialect heaven.



Evjen, L. R. 2011. æ sitt med klump i halsen når æ skriv det hær: dialekt i skriftspråket i debattforum knytt til tre norske nettaviser. MA-thesis. UiT.

Johnstone, B. & Kiesling, S.F. 2008. Indexicality and experience: Exploring the meanings of /aw/-monophthongization in Pittsburgh. Journal of Sociolinguistics 12/1, 5–33.

Rotevatn, A. G. 2014. Språk i spagaten. Facebook-språket. Om normert språk og dialekt blant vestlandselevar. MA-thesis, University College Volda.

Røyneland, U. 2018. Virtually Norwegian: Negotiating language and identity on YouTube. In Cutler & Røyneland (eds.) Multilingual Youth Practices in Computer Mediated Communication. Cambridge University Press, 145–168.

Trudgill, P. 2002.  Sociolinguistic Variation and Change. Edinburgh: Edinburgh University Press.

“I'll give them some peasanty swearing!”: Dialect and language ideologies in Greek Cypriot social media.


In this paper I examine and compare instances of use of Cypriot Greek and of alternation between Standard and Cypriot Greek in two facebook groups, Δέφτερη Ανάγνωση-Second Reading and Ομάδα Πρωτοβουλίας για την προάσπιση του Δημόσιου Σχολείου [Initiative Group for safeguarding public education]. Both groups have members in the region of 10,000 each, and both are heavily politicized: Ανάγνωση-Second Reading states as its purpose a critical take on news as presented in mainstream media, while Ομάδα Πρωτοβουλίας [Initiative Group], whose members are mainly schoolteachers, was formed recently in reaction to government measures against public education. As can be seen from the content of the posts and comments, members are typically highly literate and competent not only in Standard Greek but also in English, and they draw upon these resources, but they also deploy Cypriot Greek, their native variety.

Standard Greek, the H variety in Cyprus’ diglossic context, is the language of literacy and it is typically associated with status, overt prestige and modernity, but also with artificiality, while Cypriot Greek is reserved mostly for private oral communication, and is frequently labelled horkátika ‘peasanty’; the dialect is associated with ‘authenticity’, solidarity and local identity (Hadjioannou et al., 2011; Tsiplakou, 2011).

Recent studies of dialect use in CMC (e.g. Themistocleous, 2015) show that switching between standard and dialect is frequently capitalized on for discourse purposes and also that Cypriot Greek may be used performatively to index humor, playfulness etc. In our case, a quantitative analysis of posts and comments by focal users (i.e. members who are particularly active in the groups) suggests that Standard Greek is dominant but that dialect is also ubiquitously present. Qualitative analysis of the distribution of the two varieties indicates that Standard Greek is typically reserved for the transmission of factual information and for argumentation, while the dialect is reserved for personal, emotive commentary. Language use in these social media groups then arguably reflects mainstream language ideologies: dialect is claiming space and visibility in the public sphere, but the standard is still dominant in content associated with ‘precision’, ‘cohesion’, ‘emotional detachment’, etc. In other words, the standard still indexes literate identities while the dialect indexes ‘authenticity’ of affect (Coupland, 2014), thereby only marginally upsetting the indexical order (Silverstein, 2003).



Coupland, N. 2014. Language, society and authenticity: Themes and perspectives. In V. Lacoste et al. (eds.), Indexing Authenticity: Sociolinguistic Perspectives, 14-40. Berlin: de Gruyter.

Hadjioannou, X., S. Tsiplakou & M. Kappler 2011. Language policy and language planning in Cyprus. Current Issues in Language Planning 12, 503-569.

Silverstein, M. 2003. Indexical order and the dialectics of sociolinguistic life. Language and Communication 23, 193-229.

Themistocleous, C. 2015. Digital code-switching between Cypriot and Standard Greek: Performance and identity play online. International Journal of Bilingualism 19: 282-297.

Tsiplakou, S. 2011. Linguistic attitudes and emerging hyperdialectism in a diglossic setting: young Cypriot Greeks on their language. In C. Yoquelet (ed.), Berkeley Linguistic Society 29. Special Volume: Minority and Diasporic Languages of Europe, 120-132. Berkeley, CA: University of California at Berkeley.

Italo-Romance dialects online: functions and features of Piedmontese in social media dialogues


One of the key elements of the sociolinguistic situation characterizing Italy is the coexistence of Italian and Italo-Romance dialects, the latter being spoken across the country since the Middle Ages. Italo-Romance dialects are ‘primary dialects’ (in Coseriu’s terminology) which, in most areas, can still be deemed Abstand languages (see e.g. Maiden & Parry 1997); they can be considered as unofficial regional languages, functionally subordinated to Italian. The use of such languages is generally restricted to informal situations, especially family domains, and is clearly decreasing from generation to generation. However, partly due to a recent change of attitude towards Italo-Romance dialects, the latter can be found as well in less expected domains, even among young speakers (see e.g. Dal Negro & Vietti 2011). Such is the case with digitally-mediated communication.

This paper deals with the use of Italo-Romance dialects in the digital sphere, with particular reference to social media platforms. A general overview of such issue will be provided, drawing upon the results of previous studies (see e.g. Fiorentino 2006; Marcato 2015). The focus will then turn to the online use of Piedmontese, i.e. an Italo-Romance dialect spoken in the northern Italian region of Piedmont, which can be seen as a case in point. My arguments will draw primarily on a collection of dialogical texts written in Piedmontese, extracted from the itWaC Corpus (https://www.sketchengine.eu/itwac-italian-corpus); such data will be supplemented by a selection of dialogues from the Facebook group Nui parluma piemunteis, “we speak Piedmontese” (https://www.facebook.com/groups/nuiparlumapiemunteis).

Specific attention will be given to the purposes for which Piedmontese is used in online conversations. Reference will be made to the following four functions, which can be ascribed to the use of Italo-Romance dialects in contemporary Italy (Berruto 2006: 120): (i) everyday conversation (highest vitality), (ii) playful/expressive functions, (iii) symbolic/ideological values, (iv) folkloric uses (lowest vitality). At the same time, the linguistic, orthographic, and stylistic choices of Piedmontese speakers will be investigated. In fact, as with many other Italo-Romance dialects, Piedmontese has a certain degree of codification, and a sort of ‘prescriptive correctness’ could be expected on some platforms for ideological reasons (see e.g. Miola 2013).

The results of this study will be discussed against the backdrop of Auer (2005)’s typology of European dialect/standard constellations and in connection with the UNESCO Language Endangerment Framework.



Berruto, G. (2006). “Quale dialetto per l’Italia nel Duemila? Aspetti dell’italianizzazione e risorgenze dialettali in Piemonte (e oltre)”. In Sobrero, A. A. & Miglietta, A. (eds.), Lingua e dialetto nell’Italia del Duemila, Galatina: Congedo, 101-136.

Dal Negro, S. & Vietti, A. (2011). “Italian and Italo-Romance dialects”. International Journal of the Sociology of Language 210, 71-92.

Fiorentino, G. (2006). “Dialetti in rete”. Rivista italiana di dialettologia 29, 111-147.

Maiden, M. & Parry. M. (eds.) (1997). The dialects of Italy. London: Routledge.

Marcato, G. (ed.) (2015). Dialetto. Parlato, scritto, trasmesso. Padova: Cleup.

Miola, E. (2013). “A Sociolinguistic Account of WikiPiedmontese and WikiLombard”. Sociolinguistica 27, 116-131.

Flemish online teenage talk: oral and digital vernacular and their social correlates


Linguistic practices in written informal computer-mediated communication (CMC) challenge the operationalisation of standard and especially non-standard language, not only because CMC led to a “pluralisation” and “localisation” of written language norms, but also because it requires the inclusion of a linguistic level that until recently had largely remained beyond the scope of sociolinguistic research, i.e. spelling and “visual variability” (Androutsopoulos 2011: 151-155).  Informal CMC deviates from formal standard writing both through the integration of substandard spoken language features (dialect, regiolect, social slang) and through the presence of typical characteristics of online writing. Therefore, following  Androutsopoulos (2011: 146), we differentiate between oral (old) vernacular and digital (new) vernacular.

We will discuss social variation patterns in informal online communication on social media like WhatsApp and Facebook. The corpus consists of more than 2.5 million tokens and contains conversations produced in 2015 and 2016 by 1384 Flemish high school students aged 13-20. The linguistic variables in our research include a wide range of ‘non-standard’ markers and the social variables comprise age, gender and education (and by extension also parental profession). The present research fits in with quantitative correlational sociolinguistic studies with a primary focus on the interaction of social variables and with studies on adolescent speech that take into account age grading patterns (see e.g. Eisikovits 2006 for a combination of both).

While ‘non-standardness’ clearly correlates with all of these social variables, the analyses show that, first of all, the social variables should not be analysed in isolation, and, second, the set of non-standard markers needs to be declustered into several subsets. With respect to the first finding, age, gender and social background appear to correlate systematically.  As to the second finding: the distinction between oral and digital vernacular appears to be highly relevant. It allows for a nuanced perspective on the linguistic behaviour of different social groups.  E.g. with respect to working class youngsters’ online practices: while we expect this group to surpass their peers in the use of traditional vernacular, the data show that they also maximally exploit the potential of digital culture for personal or group identity construction.

We will end with a critical note on the challenging operationalization of oral vernacular, since automatic data extraction inevitably creates noise and prevents more subtle differentiation between e.g. local dialect features and general colloquialisms.



Androutsopoulos, J. (2011): Language change and digital media.  A review of conceptions and evidence.  In Kristiansen, T. & N. Coupland (eds.), Standard Languages and Language Standards in a Changing Europe: 145-159. Oslo: Novus Press

Eisikovits, E. (2006): Girl-talk/Boy-talk: Sex differences in adolescent speech.  In Coates, J. (ed.), Language and gender. A reader: 42-54.  Oxford: Blackwell.

12:25-13:30Lunch Break
13:30-15:10 Session 3A: Syntactic Variation
Location: Brussels
On the acceptance of grammatical cases of doubt in German


To express certain concepts speakers and writers of German may dispose of different variants. There are for example some prepositions that vary as to the case they govern. The prepositions trotz (‚despite‘) and wegen (‚due to‘) for instance may govern the genitive (trotz des Wetters, wegen des Wetters) or the dative (trotz dem Wetter, wegen dem Wetter). This variation can i.a. be described in perspective of morphosyntactic language change and the status of the variants according to their standardness (e.g. Eichinger & Rothe 2014, Eichinger 2017). Taking yet another perspective, one can focus on the acceptance of these variants – especially as they represent very prominent instances of grammatical so called cases of doubt. As such they play a dominant role in discourses on correct language. For several years now, linguistic research has been focussing on the perception of phenomena, e.g. in dialectology and in sociolinguistics. This focus can also be found in research on grammatical variation, where it is certainly important to differentiate between attitudes of laymen and those of professionals, e.g. teachers. There are e.g. studies on teachers’ ratings and handling of these variants (e.g. Davies 2000ff.), studies on their acceptance in the spoken standard variety (Schneider et al. 2018), as well as studies on doubts speakers have about these variants (e.g. Szczepaniak & Vieregge 2017).

In our paper, we will present what people in Germany think about some of these variants. We will use brand new results of a national and representative survey, i.e. the Germany Survey 2017/2018 (cf. Adler & Plewnia 2018). This survey collected data i.a. on language repertoire and language attitudes of 4,339 German residents. The depth of the data allows for discrimination of the raters according to different aspects, for example sociodemographic factors as age and level of education – thus, a differentiation of proper laymen and linguistic professionals is equally possible, Respondents can also be differentiated according to other language attitudes. In conclusion, we will compare the findings of our representative survey to those of former studies on the subject.



Adler, A. & A. Plewnia (2018): Möglichkeiten und Grenzen der quantitativen Spracheinstellungsforschung. In Lenz, A. N. & A. Plewnia (Hrsg.): Variation – Normen – Identitäten: 63-98. Berlin/Boston: de Gruyter.

Davies, W. V. (2000): Linguistic norms at school: a survey of secondary-school teachers and trainee teachers in a central German dialect area. ZDL 67: 129-147.

Eichinger, L. M. (2017). Gesprochene Alltagssprache. In Deutsche Akademie für Sprache und Dichtung/Union der deutschen Akademien der Wissenschaften (Hrsg.): Vielfalt und Einheit der deutschen Sprache: 283-331. Tübingen: Stauffenburg.

Eichinger, L. M. & A. Rothe (2014). Der Fall der Fälle. Entwicklungen in der nominalen Morphologie. In Plewnia, A. & A. Witt (Hrsg.): Sprachverfall? Dynamik – Wandel – Variation: 71-97. Berlin/Boston: de Gruyter.

Schneider, J. et al. (2018) Gesprochener Standard in syntaktischer Perspektive. Tübingen: Stauffenburg.

Szczepaniak, R. & A. Vieregge (2017): (mit) Sprache zu untersuchen, ist dank des Internets/dem Internet heute viel einfacher – Kasuswahl nach Präpositionen. Praxis Deutsch: 42-46.

Embedded V2 in Faroese and Mainland Scandinavian: results from a production study


The loss of generalized V2 in embedded clauses in Mainland Scandinavian (Msc) is a well-studied phenomenon and has been claimed to coincide with loss of rich inflection (Vikner 1995 etc.). The MSc languages still allow embedded V2 (EV2), but only in cases where the complement is (or could be) assertive (Wiklund et al. 2007). The standard analysis is that MSc has lost so-called V-to-I movement, and only utilises V-to-C movement in embedded clauses; as diagnosed by topicalization in all types of embedded clauses that allow verb movement over a sentence adverbial. As shown in Jonas (1996) and Heycock et al. (2010), Faroese has moved in the same direction, and now has a pattern where verb movement over a sentence adverbial is primarily possible in contexts where topicalization is also allowed. Results from a magnitude estimation test in Heycock et al. supported this claim by showing no quantifiable difference in the acceptance of EV2 between Faroese and Danish.

In this talk, we present the results from a large-scale production experiment carried out in the Faroe Islands, Norway, Sweden and Denmark. In the experiment, speakers were asked to turn main clauses into embedded clauses (“I always go to bed early.” à “Sam says that…”). Three types of embedded contexts were included (12 examples each): complements of “say” (assertive/bridge), complements of “be proud of” (factive), and embedded yes/no questions.

Our results support the previous findings: EV2 in Faroese is significantly more common after a verb that selects for an assertive complement than in a non-assertive complement or an embedded question. Verb movement under “say” occurred in 33% of times (and 30 out of 33 speakers (90%) produced at least one EV2 under “say”). EV2 under questions was scarce, and restricted to a small number of individuals. Verb movement under factive verbs was more common but similarly restricted. However, although the results show that Faroese has developed a system where verb movement in embedded clauses is constrained by similar semantic and information structural factors as in MSc, our results from Norway, Sweden and Denmark look quantitatively very different: EV2 in these locations was only rarely produced in the elicited sentences (yet corpus studies show them to be quite common), and only very few speakers (10-25%) ever produced them. We will argue that Faroese has developed similar restrictions for embedded verb movement (only possible in certain types of complements), whereas certain associations between word order and pragmatics/semantics have developed in MSc (as in Hopper’s “Specialization”).



Heycock, C. et al. (2010): V-to-I and V2 in subordinate clauses: an investigation of Faroese in relation to Icelandic and Danish. Journal of Comparative German Linguistics 13: 61–97.

Jonas, D. (1996). Clause structure and verb syntax in Scandinavian and English. PhD thesis, Harvard


Vikner, S. (1995). Verb movement and expletive subjects in the Germanic languages. Oxford: Oxford

University Press.

Wiklund, A.L. et al. (2007): Rethinking Scandinavian verb movement. Journal of Comparative German Linguistics 13: 203-233.

Dynamics of Gender Marking in Dutch and North-Brabantish


This paper is on changes in marking gender of nouns. Where Standard Dutch has two types of gender (common and neutral), dialects in North Brabant in the south of The Netherlands have three (masculine, feminine, neutral; Hoppenbrouwers 1983, De Schutter 2013). Gender markers belong to the most prominent features of these dialects. With a process of dialect change well on its way for at least fifty years, the knowledge of lexical gender is supposedly fading away. Also in the light of general tendencies of deflection in Germanic languages, gender marking might be expected to vanish from the dialects.

The situation, however, is completely different. In all generations of speakers in North Brabant, we find a high level of heterogeneity when it comes to gender marking. On the one hand we find dialect levelling where gender marking is disappearing in certain contexts, but on the other we find gender marking where it should not appear, according to the traditional rules of dialect grammar. An over-generalization of the application of markers leads to hyperdialectisms (Lenz 2004). Furthermore, gender is also marked by innovative constructions, e.g. in accumulate forms with two, sometimes identical suffixes (Doreleijers 2017).

This project aims to explain this phenomenon of dynamic gender marking. Our data consist of online speech (vlogs), translation tests, grammaticality assessments, and interviews on attitudes and prestige. Our project seeks to combine the linguistic variables or conditions for variation in gender marking, with sociolinguistic processes. We hypothesize that variation patterns are triggered by language internal factors (such as a general linguistic process of deflection, cf. Audring 2006) that go together with stylistic and social/cultural orientations. Primary dialect features, such as gender markers, are part of a speech style, indexing regional identities (Johnstone & Kiesling 2008). Seemingly, if one identifies with Brabant, one must sound Brabantish.

This paper will discuss the ‘what, how and why’ of the dynamics and diversity and the underlying linguistic systematics and sociolinguistic processes, as well as how the case of North Brabant can tell us more on changes of marking gender in general (cf. Tamminga 2013).



Audring, J. (2006) Pronominal gender in Dutch. Journal of Germanic Linguistics 18, 85-116.

De Schutter, G. (2013) The dialects of Brabant. Grammatical properties. Language and Space: An International Handbook of Linguistic Variation, Volume III: Dutch, ed. F. Hinskens & J. Taeldeman, 297-318. Berlin/New York: Mouton de Gruyter.

Doreleijers, K. (2017) ‘Unne kuukske’? Een onderzoek naar variatie in geslachtsmarkering door (jonge) Brabantse dialect-/regiolectsprekers in regio Eindhoven. Master thesis Neerlandistiek, Universiteit Utrecht.

Hoppenbrouwers, C. (1983) Het genus in een Brabants regiolect. Tabu 13, 1-25.

Johnstone, B. & S.F. Kiesling (2008) Indexicality and experience: Exploring the meanings of /aw/-monophthongization in Pittsburgh. Journal of Sociolinguistics 12, 5-33.

Lenz, A. (2004) Hyperforms and Variety Barriers. Language variation in Europe. Uppsala University, 281-293.

Tamminga, M. (2013). Phonology and morphology in Dutch indefinite determiner syncretism: Spatial and quantitative perspectives. Journal of Linguistic Geography 1(2), 115-124.

13:30-15:10 Session 3B: Sociophonetics
Location: London
The role of ethnolects in variation and change: Vowels in Sydney English

ABSTRACT. With increasing migration across the globe, there has been an upsurge of interest in “ethnolects”, varieties of speech used by ethnically diverse groups of young people in modern metropolises (e.g., Cheshire et al. 2011; Gross et al. 2016; Hoffman & Walker 2010; Wiese 2009). A key question that arises in such work concerns the role of ethnic communities in variation and change, including their participation in, and impact on, patterns in the wider speech community.

In Australia, we are in a unique position to address this question, thanks to foundational work by Barbara Horvath in the 1970s,analysing the speech of Anglo-, Italian- and Greek-Australians (Horvath 1985). Here, we re-examine a subset of these 1970s data, for comparison with a corresponding corpus of sociolinguistic interviews recorded in the 2010s (Travis 2016-2021). We focus, for both the 1970s and 2010s, on Anglo- and Italian-Australians (a well-established migrant group with a long presence in Australia), and, for the 2010s, on Chinese-Australians (a newer migrant group). The combined data comprise over half a million words of spontaneous speech from 140 AustraliansYounger and older speakers are captured in each time period (with birth dates ranging from the 1930s to the 1990s), allowing for change to be tracked in real and apparent time.

We conduct acoustic analyses of over 20,000 tokens of a set of vowels considered to be characteristic of Australian English: fleece, face, goat, price and mouth. Consistent with prior reports (Cox & Palethorpe 2012), we find change over time for these vowels, which can be generally described as movement away from canonical, working class, “broad” Australian realisations. We further observe that these changes were forecast in the 1970s by middle-class Anglo female teenagers, and by Italian-Australians. A gender effect also surfaces, where men have lagged behind women over time, such that they tend to retain broader realisations than women, an effect that holds for both Anglo- and Italian-Australians. In addition, while we see a robust socio-economic class distinction for Anglos in 1970s speakers, particularly for men, these class distinctions have attenuated over time. Finally, in the 2010s, Chinese-Australians exhibit the least broad realisations in comparison with both Anglo- and Italian-Australians of comparable socio-economic class.

These findings indicate that migrant groups do indeed participate in changes taking place in the broader society, and they may help to drive these changes forward.


Cheshire, Kerswill, Fox & Torgersen. 2011. Contact, the feature pool and the speech community. Journal of Sociolinguistics 15(2): 151-196.

Cox & Palethorpe. 2012. Standard Australian English: The sociostylistic broadness continuum. In Hickey (ed.), Standards of English, 294-317. Cambridge: CUP.

Gross, Boyd, Leinonen & Walker. 2016. A tale of two cities (and one vowel). Language Variation and Change 28(2): 225-247.

Hoffman & Walker. 2010. Ethnolects and the city. Language Variation and Change 22(1): 37-67.

Horvath. 1985. Variation in Australian English. Cambridge: CUP.

Travis. 2016-2021. Sydney Speaks. ARC Centre of Excellence for the Dynamics of Language, ANU: http://www.dynamicsoflanguage.edu.au/sydney-speaks/.

Wiese. 2009. Grammatical innovation in multiethnic urban Europe. Lingua 119(5): 782-806.

The effect of language contact and gender on the vowel system of Mišótika Cappadocian


The aim of this paper is to present the changes in the vowel system of contemporary Mišótika, the variety of Cappadocian Greek originally spoken in the village Mistí, in what is now the Central Anatolian Region of present-day Turkey.

Cappadocian is a linguistic variety of Greek origin which had been in contact with Turkish for almost nine centuries until the population exchange between Greece and Turkey (1924) when the Cappadocians were forced to settle in Greece.  Before the population exchange, the Cappadocian vowel system consisted of eight vowels, aligning it with the vowel system of Turkish, i.e. /a, e, i, o, u, œ, y, ɯ/ (Dawkins 1916; Janse 2009).

The present study is an acoustic analysis of vowels of Mišótika and it is based on recordings of 20 native speakers, who live in two different Cappadocian communities in Northern Greece, one in the prefecture of Kilkis (Neo Agioneri), and the other in Thessaloniki (Xirohori). The first one is a homogeneous village, whereas the second is a mixed village, inhabited not only by Cappadocians but also by other Greek-dialect speakers. In particular, we compare the speech of Mistiot speakers from the two communities and analyse the differences between them, taking into consideration mechanisms of language contact and linguistic change (see Chambers & Schilling (2013) for a comprehensive overview). We also study the role of the sociolinguistic variable of gender, since differences are often observed between male and female speech (see Holmes & Meyerhoff (2003) for a comprehensive overview).

According to our findings, the current vowel system of Mišótika seems to diverge significantly from the older one. Specifically, three vowels which do not exist in the Modern Greek vowel system have either disappeared (/œ, y/) or are in the last stage of elimination (/ɯ/), due to contact with Standard Modern Greek and local Greek varieties (through a levelling process). Remarkably, a previously unrecorded vowel /æ/ that does not exist either in Modern Greek or in Turkish, is now prominently present in words of both Greek and Turkish origin.

At the same time, the results indicate that there are further differences, in relation to the vowel system, between the two Cappadocian speech communities, differences that are the result of contact under different sociolinguistic conditions. Finally, gender also seems to be a significant sociolinguistic parameter influencing the distribution of the vowels in the vowel spectrum, because the preliminary analysis of our data points out that male speakers are one step ahead in the process of linguistic change.



Chambers, J.K. & N. Schilling (eds.) (2013). The Handbook of Language Variation and Change (2nd edition). Oxford: Blackwell.

Dawkins, R. M. (1916) Modern Greek in Asia Minor: a Study of the Dialects of Sílli, Cappadocia and Phárasa with Grammar, Texts, Translations and Glossary. Cambridge: Cambridge University Press.

Holmes, J. & M. Meyerhoff (eds.) (2003). The Handbook of Language and Gender. Oxford: Blackwell.

Janse, M. (2009): Greek-Turkish Language Contact in Asia Minor. Études Helléniques / Hellenic Studies 17: 37-54.

Leading with style: the real time incrementation of different types of sound change


When children are included in apparent time studies of language change, they are shown to lag behind the leading adolescents. This creates what Labov (2001) terms an ‘adolescent peak’, a finding demonstrated across a range of varieties and across changes operating at different levels in the grammar (Tagliamonte & D’Arcy, 2009; Holmes-Elliott, 2016). In order for language change to continue, children must go through a process of incrementation where they increase their rates until they become the leaders.

The process is somewhat complicated by different types of sound change: those which operate above the level of consciousness, and those which speakers are largely unaware of (e.g. Labov, 1966; L. Milroy, 2007). Within adult communities, the divergent social and linguistic patterning of these types of changes is well attested. However, much less is known about how these patterns emerge during childhood. Previous research suggests that while young children acquire an ability to match style shifting patterns early on (Foulkes et al., 2005; Smith et al., 2007), the ability to internalise and exhibit adult-like patterns of style shifting does not develop until later (Kerswill & Williams, 2000). So how does the incrementation of a socially marked feature compare to one that flies below the radar?

To address this question and investigate the process directly, this paper contributes an analysis of real time data. The sample consists of 13 speakers, and targets a key phase in development – childhood to adolescence. The speakers were initially interviewed aged 9-11, and then again 4 years later, aged 13-15. An age stratified adult corpus of speech from the same community forms a baseline for comparison. Crucially, the study investigates two different types of sound change. The first feature under investigation is GOOSE-fronting, where the /u/ vowel is shifting forward in the vowel space. This feature is phonetically gradual change and typically escapes social commentary. This is in contrast to the second feature, TH-fronting, where the dental fricatives are variably replaced by labiodentals (e.g. think realised as fink). This feature is phonetically abrupt, heavily stigmatised and often commented upon.

Analysis of 5547 tokens of the GOOSE vowel, and 3720 TH-fronting contexts reveals that they do indeed show very different trajectories. For GOOSE fronting, individuals increase and converge as a population. As expected, the change increments at the individual and the aggregate level and the change shows consistent patterning across the two time points. For GOOSE-fronting, speakers flock unconsciously towards the vanguard in a linear fashion. For TH-fronting, in contrast, individuals withdraw from the change, presumably, as they develop a more nuanced understanding of its stigma. Inspection of the social and linguistic constraints of this change suggests that this withdrawal is socially motivated. Taken together, these results illustrate the intersection two developmental processes in real time: the incrementation of sound change, and the acquisition of sociolinguistic competence. More broadly, how linguistic, social and developmental factors interact to constrain and propel different types of sound change.

Rhotics degemination in Roman Italian: between production and perception


Rhotics degemination is considered a sociolinguistic marker of the Roman Italian variety, usually associated with rudeness, low status speakers and informality (Stefinlongo 1999). Given the particular status of the R variable (Scobbie 2006) in conveying socio-indexical information, we analyse if, and how, young Roman speakers produce degemination of the rhotics in Roman Italian. A perceptive test also aimed at confirming the presence and diffusion of the stereotypes among Italian native speakers of different areas. To our knowledge, no explicit sociophonetic study has been devoted to the investigation of rhotic degemination and how it is phonetically realized (but see Marotta 2005).

We recorded 10 speakers from Rome, 5 males and 5 females, aged between 23 to 30 and with high degree of education (high school diploma or PhD degree). Speakers participated to a sentence-reading task, with 70 sentences of equal length and controlled prosodic contour. Each sentence contained one token with a singleton and/or geminate /r/, in stressed and/or unstressed condition, with or without a lexical competitor. The 700 tokens were annotated in PRAAT following the protocol for rhotics developed for Sicilian Italian by Celata et al. 2016.

We then selected 12 tokens for a perceptive test involving both word recognition and a possible linguistic stereotype associated to the speaker. The test was submitted through the web to different Italian mother speakers, equally divided between Northern and Roman speakers.

Results show that, for rhotic duration, Gemination and Lexical competitor were statistically significant, since geminated rhotics were longer than singleton rhotics, and word with a minimal pair neighbour were significantly longer than words without a corresponding minimal pair. Sex and Syllable stress were not significant, but they interact in stressed syllables. The interaction between Gemination and Lexical Competitor showed that tokens with a minimal pair neighbour maximized the contrast between singleton and geminated rhotics, whereas tokens with no minimal pair neighbour tend to show a less sharp contrast between singleton and geminates. As for vowel duration, only Syllable stress was statistically significant, with stressed syllable being longer than unstressed one. Unexpectedly, no effect of Gemination was found.

Qualitative analysis of the distribution of rhotics across the two conditions, namely intervocalic singletons and geminates, show, instead, a more complex picture. Intervocalic geminate rhotics seem to allow a greater range of possibilities: they can indeed be realized mainly as a trill, but approximant, fricatives, taps, and combined realizations (trill or tap with a fricative appendix) are found too. However, a great within-speaker variety has also been observed. The perceptive test confirms that the degemination is salient independently from speakers’ origins. However, no clear correlation between degemination and linguistic stereotypes has been found.

Thus, our results confirms that rhotic degemination is still present in Roman Italian. From a sociophonetic perspective this could be interpreted as the maintenance of a marked feature to show affiliation to a cultural-defined group, but the presence of a linguistic stereotype has to be fully and more carefully investigated.



Celata, C., Meluzzi, C. and Ricci, I. (2016), The sociophonetics of rhotic variation in Sicilian dialects and Sicilian Italian: corpus, methodology and first results, in «Loquens», 3,1.

Marotta, G. (2005), Il consonantismo romano. Processi fonologici e aspetti acustici, in Albano Leoni, F. and Giordano, R. (eds.), Italiano parlato. Analisi di un dialogo, Napoli, Liguori, pp. 1-24.

Scobbie, J. (2006). (R) as a variable. In Brown, K. (Ed.), Encyclopedia of Language & Linguistics (2nd ed.), Oxford, Elsevier, pp. 337–344.

Stefinlongo, A. (1999) “Neoromanizzazione” del territorio. La lingua delle scritte murali nell'area metropolitana romana, in Dardano, M. et al. (a cura di), Roma e il suo territorio. Lingua, dialetto, società, Roma, Bulzoni, pp. 267-285.

13:30-15:10 Session 3C: Panel Coherence
Location: Madrid
Co-occurrence of neo-standard features in spoken Italian: a corpus-based study


In many parts of Europe, the well-known social changes which have taken place in late modernity have brought about the massive spread of the standard language (often to the detriment of primary dialects) and the subsequent development of a new standard variety, the latter being closer to the spoken language than the traditional, literary, one. In Italy, the twentieth century spread of the standard language has led many spoken informal features to be used and accepted even in formal and educated speech, as well as partly in formal and educated writing. At least since the 1980s, it has been claimed that a new standard variety of Italian has thus emerged, i.e. the so-called ‘neo-standard Italian’ (cf. Berruto 2012). However, no study so far has empirically tested the co-occurrence of features of this variety.

The paper will investigate the extent to which clusters of neo-standard features co-occur in speech and, therefore, whether this new standard variety of Italian can be characterized as displaying coherence. In the same vein as Guy & Hinskens (2016), our study will explore the following research questions:

a)    is neo-standard Italian made up of features which actually cluster in usage?

b)    is there greater coherence between features within the same subsystem of the grammar?

c)    to what extent might these possible differences relate to language-internal factors and/or social factors?

d)    do particular groups of speakers (e.g. poorly educated vs. highly educated individuals, as well as younger vs. older generations) behave more or less coherently depending on their social properties?

To this end, a comparison will be drawn between two datasets extracted from:

1)    [ki'parla] (http://www.leadhoc.org/index.php/data-access/corpus-of-spoken-italian): a corpus of roughly 25 hours of spoken Italian collecting several types of interaction recorded at the Universities of Torino and Bologna (academic lessons; professor-student interactions during office hours and oral examinations; within-group interactions between students or professors in spontaneous contexts; etc.);

2)    ParVa (http://www.mediling.eu): a corpus of 15 hours of semi-structured interviews with a group of former World War II partisans living in Val Camonica (Brescia); all the partisans were born in the 1920s and most of them were factory workers, peasants or mine workers, coming from low-income households.

The main aim of the comparison is to observe the distribution of a set of morphological, syntactical and lexical neo-standard features across two macro-social dimensions, namely education (highly educated vs. school educated or uneducated) and communicative situation (formal vs. informal situations, only in [ki'parla]).

The methods of corpus linguistics and variationist sociolinguistics will be combined to extract the linguistic features. A principal component analysis was performed on the dataset to identify the dimensions of variation of the linguistic features. Then a hierarchical cluster analysis was carried out to aggregate the similar linguistic features in systems of coherent variables.



Berruto G. (2012), Sociolinguistica dell’italiano contemporaneo. 2nd edn. Roma, Carocci. (1987: Roma, La Nuova Italia Scientifica).

Guy G. & Hinskens F. (2016), “Linguistic coherence: Systems, repertoires and speech communities”. Lingua 172-173: 1-9.

Coherence in ongoing varieties. The effect of mesosocial and small-scale variables on the use of the intermediate (standard/vernacular) variety in southern Spain.


This paper addresses some questions regarding coherence within ongoing urban varieties, such as the one boosted by urban middle-class speakers from southern Spain since the fifties (Villena and Vida 2017, 2018). The main issues we focus on here are (1) whether coherence actually exists within a variety which has primarily emerged on the basis of correction and levelling of vernacular features, and (2) to what extent it is feasible to build a multivariate model able to depict interaction between the macrosocial, mesosocial and small-scales variables underlying the speaker’s use of the new variety.

Two relevant factors seem then to be decisive:

(1) On the one hand, the existence of a social and ideological change capable of producing blending of different and even contradictory linguistic features. The role of a proactive speaker willing to use the new variety is then essential.

(2) On the other hand, although coherent varieties are composed of many single variables which covary (Guy and Hinskens 2016), it seems that speakers identify these variables with a particular, most prominent, feature which defines itself as the nucleus of the variety around which all the other variables turn. This is the case with the southern demerging of dental fricative /θ/.

This presentation deals with coherence within an emerging intermediate (standard / vernacular) interregional variety shaped as a consequence of a new identity involving both traditional regional/local and modern/national  identities. The social meanings involving this mixed identity are conveyed through combination of vernacular and standard speech features.

The analysis of a certain number of features characterising the Andalusia vernacular varieties has revealed that most of them have undergone levelling and been replaced by standard-like variants among urban middle-class speakers. Distribution of these variants points to the emergence of a new variety which symbolises proactive integration into modern life and the national well-being society. A multivariate SEM model (path analysis) shows that mesosocial and small-scale variables reflecting the speaker’s linguistic attitudes and orientation towards the standard act as mediators between macrosocial entities and the individual speaker’s speech behaviour (Villena-Ponsoda 2018).



Guy, Gregory and Frans Hinskens (eds.) (2016). Linguistic coherence: Systems, repertoires and speech communities. Lingua 172-173.

Villena-Ponsoda, Juan-Andrés (2018). The dilemma of the reliability of geolinguistic and dialectological data for sociolinguistic research. The case of the Andalusian demerger of /θ/. 9th Congress of the International Society for Dialectology and Geolinguistics (SIDG 2018). Vilnius, 23-27 July, 2018.

Villena-Ponsoda, Juan-Andrés and Matilde Vida-Castro (2017). Between local and standard varieties: horizontal and vertical convergence and divergence of dialects in Southern Spain, In Isabelle Buchstaller y Beat Siebenhaar (eds.): Language Variation. European Perspectives VI, Amsterdam: John Benjamins, 125-140.

Villena-Ponsoda, Juan-Andrés and Matilde Vida-Castro (2018). Variation, identity and indexicality in southern Spain. On the emergence of a new intermediate variety in urban Andalusia. In Stavroula Tsiplakou and Massimo Cerruti (eds.), forthcoming.

Variation and change in Afrikaans dialects


From the 19th century onwards the Afrikaans verbal system was characterised, inter alia, by the loss of the infinitive, the loss of the preterite as a verbal category, the regularisation of the past participle and the rise of verb clustering. In this paper variation will be compared in the verb systems of the three main historical dialects of Afrikaans, namely (A)  Eastern Border Afrikaans (the basis of the present standard), (B) Cape Afrikaans and (C) Orange River Afrikaans.  The following variables will be discussed in broad outline:

Past participle formation: Afrikaans past participles are characterised by a rising stress pattern, e.g. gegáán and probéér, in the absence of which the prefix ge- is obligatory, e.g. gehárdloop, rendering ge- optional in many cases. At present we find a marked difference between lect A and lect C in that lect A drops optional ge- whenever possible (e.g. probéér, verslááp), while lect C employs ge- profusely (e.g. geprobeer, geverslaap).

Preterite retention: After the loss of the preterite as a tense category the modal preterites sou ‘would’, wou ‘wanted to’, kon ‘could’ and moes ‘had to’ constituted almost the full set of extant preterites. Two constructions are available for these modals to express past tense, viz. modal preterite + base form of main verb, as in kon sing, lit. ‘could sing’, or modal present + past participle of main verb + het, as in kan gesing het, lit. ‘can have sung’. When no preterite is extant, only the second option is available, e.g. mag gesing het, lit. ‘may have sung’. We find that lect A has a marked preference for the first construction, where available, while the second option is more often employed in lects B and C.

Het/hê  ‘have’ and is/wees ‘be’ as auxiliary or main verb/copula:  and wees, the last bare infinitives still remaining in Afrikaans, are frequently used in lect A where they have even acquired new functions, e.g. Kom ons wees eerlik ‘let us be honest’, lit. ‘come we be honest’. However, in lect B in particular, is almost consistently replaced by het and wees often replaced by is, e.g. Wanneer sal jy klaar is? ‘When will you be finished?’, lit. ‘when will you finished is’.

Verb clustering, as in Môre begin werk hulle weer (as against Môre begin hulle weer werk) ‘Tomorrow they are starting to work again’, has become very prominent in Afrikaans and often serves to express aspect. It is noticeable that clustering is used more extensively in lect C than either A or B, e.g.  Net soos die brood kom begjeent rys in die skottel … ‘As soon as the bread starts rising in the dish’, lit. ‘Just as the bread come begin rise in the dish’.

It therefore seems that while lect A is characterised by the retention and even functional renewal of inflectional relics, lect C has a preference for forming more analytical constructions, with the position of lect B still in need of clarification.

Indexicality and Cohesion


The conventional sociolinguistic model of the speech community (Labov 1966; Gumperz 1968) views it as involving shared linguistic characteristics and norms for their usage.  The social indexicalities of the linguistic features are indicated in production by social stratification and shared patterns of style shifting, and in perception by subjective evaluations of speakers using those traits.  The indexicalities in this model are principally oriented along a single dimension of prestige (overt or covert), and are implicitly cohesive – e.g. higher status speakers should use more of all prestige variants.  Current work on indexicality, however, envisions a much broader range of indexical evaluations of linguistic features, associated with social meanings of style, stance, and identity.  This raises the question of whether or to what extent different features can be coherent, if they can potentially mean very different things.  It also problematizes the speech community model: is the community defined exclusively by the shared norms on a unidimensional prestige scale, or multidimensionally, incorporating shared evaluations of other kinds of social meanings of linguistic variables.  

This paper explores these issues using data on multiple variables occurring in São Paulo Portuguese. By their social distributions, some of these appear to have conventional prestige-linked indexicalities that are supraregional (e.g. nominal and verbal number agreement, which shows social and stylistic stratification throughout Brazil).  Others are associated with regional identity – ‘Paulistanity’; belonging to São Paulo city (tapped coda /r/ and diphthongal EN) and ‘Northeastern’: from families who migrated to SP from Northeastern Brazil (unpalatalized /t,d/ before [i], fricative /r/ and repeated negation); with ruralness (unpalatalized /t,d/, /r/ deletion, nonstandard agreement); with femininity (diphthongal EN, standard agreement), and with a change in progress (diphthongal EN).  Some therefore have polyvalent indexicality, and some have overlapping indexicalities on certain dimensions.   They also differ in speaker awareness: nonstandard agreement is a stereotype of lower class/education nationwide, but it also weakly indexes masculinity (perhaps constituting an indicator); diphthongal EN is a change from below which Paulistans show little awareness of, but which is recognized by outsiders as an SP feature. 

We explore the hypothesis that coincident indexicalities are associated with greater covariation, and more polyvalent variables may show stronger or weaker covariation depending on whether their multiple indexicalities reinforce or conflict.   In a speech sample of 40 Northeastern migrants living in SP, we find that unpalatalized /t,d/ covaries with /r/ deletion, nominal agreement and negation, with which it shares the rural-urban, educational level and North-South traits respectively. On the other hand, the realization of coda /r/ as taps or fricatives, related only to North-South indexicalities, doesn't covary with any other variable.  Among 118 speakers native to SP, /r/ deletion and all agreement variables covary, forming a cluster of variables that index class/education, while dipthongal EN is uncorrelated with any other variable, being an innovation showing little community-internal awareness.



Gumperz, John. 1968. The speech community.  International encyclopedia of the social sciences: 381-386. Macmillan. 

Labov, William.  1966.  The social stratification of English in New York City.  Washington DC: Center for Applied Linguistics.

13:30-15:10 Session 3D: Panel Dialectometry
Location: Moscow
Development of Lithuanian dialectal classification across centuries


The 17th–19th century differentiation of Lithuanian dialects is based on the administrative–territorial division of the area of the Lithuanian language. Two most distinct classifications of late 19th century are those made by Antanas Baranauskas, a Lithuanian poet and lecturer at theological seminary, and his student Kazimieras Jaunius.

They both are relevant for Lithuanian dialectology, as both of them possessed typical attributes of the Neogrammarians and were followed on, serving as a basis for the scientific 20th century dialectal classifications.

The classifications of Lithuanian dialects are grounded on the phonetical and phonological characteristics of the dialects; as a result, the Neogrammarian traditions had been kept in Lithuanian right until the end of the 20th century.

A comprehensive study of dialects done in early 21st century showed that dialects and subdialects are in flux: the locals tend to gradually drop the primary characteristic of their dialects, but preserve the secondary and tertiary attributes quite well. A clear relationship between preserving the characteristics of dialects and the age of the respondents has been observed. Which has led to a conclusion that new territorial dialectal formations that no longer have any distinct boundaries or borders are taking shape. The habitats of those regional dialects (or regiolects) roughly overlap the territories of Lithuania’s ethnographical regions.

At the beginning of the 21st century, Lithuania introduced a multifunctional model to study dialects, which targets not only the language of local residents, but their environment (sociocultural networks, linguistic landscape, and so on) and the users’ attitude towards their own language variation(s). Therefore, the field of perceptive dialectology is undergoing rapid development in Lithuania at this time. The qualities of the respondents’ language and their answers to survey questions are addressed using dialectometrical methods that clearly disclose the characteristics of linguistic variation.

The multifunctional model can be easily adapted to serve the needs of regional studies. It works especially well when a subject territory has not one, but several variants of language (or even variants of different local languages). Using this method, we can predict the organic course of the development of a language variation and its shift in the face of changing economic, demographical, or other circumstances. More details about it will be given in the report.

Dialect classification of Slovene – history and perspectives


In the paper dialect classification of the Slovene language will be presented and critically evaluated.

Theoretical framework: The knowledge of dialect diversity of Slovene language is as old as the Slovene standard language (Trubar 1555). In the 19th century many linguists have tried to classify Slovene dialects, mostly from a diachronic point of view (Kopitar 1808; Miklošič 1879; Oblak 1894; Sreznjevski 1841, etc.). In the 20th century Slovene dialectology developed as an academic linguistic discipline, and the first dialect maps of Slovene were published (Ramovš 1931) with an explanation of the methodology these dialect classifications are based on (which has mostly been so called »auditory impression« and researches of diachronic development of Proto Slavic phonemes in Slovene dialects).

The Slovenski lingvistični atlas (SLA) [Slovenian Linguistic Atlas (SLA)], which is the basic work of modern Slovenian dialectology and geolinguistics, was conceived by the linguist Fran Ramovš in 1934 – in the second half of the 20th century many publications were written based on the analysis of the SLA dialect material, together with The map of Slovene Dialects by T. Logar and J. Rigler (1983). The first two volumes of SLA have been published in 2011 and 2016 and they depict the use of the dialect vocabulary in 417 Slovenian places as recorded in the last few decades. These volumes, containing the vocabulary from the fields of "man – body, diseases, family" and "farm", provide about 4,500 dialect lexemes for 239 questions (concepts) in many dialect phonetic variants (e.g. standard Slovene grlo ‘throat’: ˈgərlə, áːrlo, hàrlo, háːrlu, gḁːʀˈlə, ˈgərlȯ, etc.). Studies of these data sets as well as newer researches of local dialects provided new insights into Slovene dialect classification in comparision with older dialect classifications (geographical differentiation, administrative divisions, cultural and ethnological influences, modern migrations etc.).

In the paper some new insights in the dialect classification of Slovene will be presented. One of the most important are its sociolinguistic aspects (from the perspective of language contact, daily migrations, the influence of the official language), especially on language borders, where Slovene gets in contact with naighbouring languages and their dialects. This will be presented on two examples: Ziljsko dialect of village Rateče at the border of Slovenia, Austria and Italy and Čabranško dialect at the Slovene-Croatian border.



Kopitar, J. (1808). Grammatic der Slavischen Sprache in Krain, Kärnten und Steyermark. Laibach: Wilhelm Heinrich Korn.

Logar, T., and Rigler, J. (1983). Karta slovenskih narečij. Ljubljana: Geodetski zavod SRS.

Miklošič, F. (1879). Vergleichende Grammatic der slavischen Sprachen. Wien: Wilhelm Braumüller.

Oblak, V. (1894). Archiv für slavische Philologie, XVI. Berlin: Weidmannsche Buchhandlung.

Ramovš, F. (1931). Dialektološka karta slovenskega jezika. Ljubljana: Rektorat Univerze Kralja Aleksandra I., J. Blasnika nasl. Univerzitetna tiskarna.

Sreznjevski, I. I. (1841). O narečijah slavjanskih. Žurnalъ ministerstva narodnago prosvješčenija. Častъ XXIX. Sanktpeterburgъ: Vъ tipografii Imperatorskoi Akademii Naukъ.

Trubar, P. (1555). Ta evangeli svetiga Matevsha, sdai pervizh vta Slouenski Iesig preobernen. Tübingen.

Identification of lexical areas templates throughout the Occitan domain.docx


The project presented here began several years ago. Its purpose is to show the relationship between the limits of lexical areas and the limits of dialectal areas traditionally defined in the Occitan domain.

Based on data from regional language atlases implemented in the Thesaurus Occitan (http://thesaurus.unice.fr/index.html), a first study (Brun-Trigaud 2013) enabled the identification of several types of areal layouts, as well as 8 templates in which a large number of Occitan lexical areas fits. This starting point had led us to question how to name the entities that do not fit in the traditional dialectal template.

On an extended database, the work was resumed (Brun-Trigaud 2017) and processed in Gabmap (https://gabmap.nl) where analyses by multidimensional scaling and Fuzzy Clustering confirmed the results of our first approach.

Within the ANR project named ECLATS (https://eclats.imag.fr/), other methods of multidimensional spatial analyses are currently experimented, allowing the exploration of the lexical data from a complementary point of view. The proposed analytical procedure first computes the various lexical areas associated to each of the lexical entries. The lexical areas are stored as spatial objects and become the statistical units of the subsequent spatial analyses, contrary to the dialectometric approach that considers survey points.

The identification of recurrent areal layouts throughout the different lexical areas is performed by computing pairwise concordance indicators that measure whether two given spatial partitions present independent spatial distributions or high spatial concordances. Several concordance indicators have been tested and compared. Using various classification algorithms, we are able to define stable clusters, each one giving out a specific layout template of lexical areas.

We also propose a new methodology of spatial analyses in order to cluster the lexical areas according to historical or geographic criteria (old provincial areas, mountain ranges, forest areas (…) and detect spatial synchrony between them. The following steps describe the process: (i) Discretization of our studied territory into a lattice of spatial units or cells (ii) Projection of these spatial units in a synthetic low dimensional representation space, using Correspondence Analysis or Multiple Correspondence Analysis (Rencher 2002). The factorial axes are spanned according to the chosen historical or geographic criteria (iii) Plotting of the lexical areas on the synthetic space as barycentric centroids (iv) Clustering the lexical barycentric centroids (v) Reconstruction of the templates using a reverse process.

Recent tools in spatial analyses of ECLATS project open new perspectives in the field of dialectology regarding the classification of lexical areas into spatial templates.



Brun-Trigaud, G. & Malfatto, A. (2013). Limites dialectales vs limites lexicales dans le domaine occitan: un impossible accord?. In Carrilho, E. (éd.): Current Approaches to Limits and Areas in Dialectology: 293-310. Cambridge Scholars Publishers.

Brun-Trigaud, G., Malfatto, A. & Sauzet, M. (2017): Essai de typologie des aires lexicales occitanes: regards dialectométriques. Fidélités et dissidences. Actes du 12e Congrès de l’Association Internationale d'Etudes Occitanes (Albi, 2017). (to be published)

Rencher, Alvin C. (2002) Methods of multivariate analysis. Wiley-Interscience, Hoboken, NJ, USA, second edition.

Internal and external language borders in the Galician domain


Given the location of the Galician in the northwestern corner of the Iberian Romance area, it has been an object of interest to dialectologists practically since dialectology began to be a scientific discipline within linguistics (Sánchez 2011). Galician belongs to a group of historical language varieties running along the northern part of the Iberian Peninsula in a dialect continuum which extends from Catalonia’s Mediterranean coast to the Atlantic seaboard (Gargallo 2011). The Galician-speaking area also forms part of the Galician-Portuguese linguistic domain, which extends from the Cantabrian Sea to the Algarve. The border between this Galician area and the Leonese linguistic domain has been studied in great detail since Menéndez Pidal founded the Hispanic linguistic school. Another issue which has attracted the interest of various scholars within a traditional dialectological approach is that of the internal divisions among Galician dialect varieties (Santamarina 1982; Fernández 1990; Álvarez, Dubert & Sousa 2006). While the separation between Galician and Astur-Leonese is viewed as a straightforward linguistic divergence in its origin, that between Portuguese and Galician is generally presented as the result of social diversification.

This contribution presents a critical review of the criteria previously used in Galician dialectology to determine its internal and external borders. Galician dialect classifications can be seen as occupying various positions on a scale between two extremes, with acknowledgment of internal variation at one end; at the other, denial of dialect divisions within Galician, maintaining the view that it is a single entity (although constituted from a cluster of speech varieties). For this purpose, we use data from several projects of geolinguistic and some methodological principles of aggregate analysis of linguistic variation (Heeringa & Nerbonne 2001; Pickl 2016; Sousa 2017).



Álvarez, R., Dubert García, F. & Sousa, X. (2006). Aplicación da análise dialectométrica aos datos do Atlas Lingüístico Galego. In Álvarez, R., Dubert García, F. & Sousa, X. (eds.): Lingua e territorio: 461-493. Santiago de Compostela: Consello da Cultura Galega: Instituto da Lingua Galega.

Fernández Rei, F. (1990) Dialectoloxía da lingua galega. Vigo: Xerais.

Heeringa, W. & Nerbonne, J. (2001): Dialect Areas and Dialect Continua. Language Variation and Change 13 (03): 375–400.

Pickl, S. (2016). Fuzzy dialect areas and prototype theory: discovering latent patterns in geolinguistic variation. In Côté, M.-H., Knooihuizen, R. & Nerbonne, J. (eds.): The Future of Dialects. Selected Papers from Methods in Dialectology XV: 75–97. Berlin: Language Science Press.

Sánchez Rei, X. M. (2011) Lingua galega e variación dialectal. Ames: Laiovento.

Santamarina, A. (1982). Dialectoloxía galega: historia e resultados. In Kremer, D. & Lorenzo, R. (eds.): Tradición, actualidade e futuro do galego. Actas do Coloquio de Tréveris: Santiago de Compostela: Xunta de Galicia – Consellería de Cultura.

Sousa, X. (2017). Aggregate analysis of lexical variation in Galician. In Buchstaller, I. & Siebenhaar, B. (eds.): Language Variation. European perspectives VI: 71-83. Amsterdam: John Benjamins.

13:30-15:10 Session 3E: Panel Minority
Location: Paris
Spanish and Nahuatl in Contact: variation in language proficiency, use and typological change


The paper discusses some of the results of 500 years of interaction between Spanish–as spoken in New Spain and then in the modern Mexican state–and Nahuatl, a Uto-Aztecan language spoken in Mexico. The corpus is comprised of several hundred colonial texts (1540s–1820s) coupled with modern data (1970s–2018) from strongly urbanized and more peripheral rural Nahua communities. The modern data come from bilingual Spanish-Nahuatl speakers who reveal varying levels of proficiency and speech behaviors in the two languages.

My point of departure is a holistic approach, correlating ethnohistorical, anthropological and social perspectives with linguistic data to pinpoint precipitating factors that are driving linguistic change in Nahuatl and Spanish. An analysis of Spanish lexical and structural influence in an extensive body of written texts in Nahuatl is complemented by present-day ethnolinguistic data from both strongly urbanized and more peripheral rural communities (Valley of Mexico, central and northern Puebla, Tlaxcala, Huastecan region in northern Veracruz). Specific examples and data include adjustments in the quantitative distributions of feature variants, (contact-induced) grammaticalization (e.g. the transformation of relational words into prepositions) and system-altering changes (e.g. the gradual disappearance of the category of animacy and the reduction of head-marking) as part of a typological change from a polysynthetic to a more analytic and fusional language, modeled on Spanish.

Focusing on present-day bilingual speakers of Nahuatl and Spanish, I discuss the results of proficiency assessment of their two languages carried out with standardized non-verbal stimuli. The proficiency assessment enabled me to perform a quantitative and qualitative comparison between different kinds of speakers coming from different regions and communities (at different stages of language shift), and also including immigrants in the US. At the next stage of analysis, I link these levels of proficiency to specific lexical, structural and pragmatic patterns and readjustments in their Spanish and Nahuatl. I looked specifically at how Nahuatl categories of animacy influence pluralization patterns in Spanish and, conversely, how Spanish pluralization of inanimate nouns impacts the usage of the indigenous language. Further insights are provided by patterns of code-switching documented with these speakers, revealing the fluidity of the matrix language and embedded language roles/choices and their dependence on the levels of proficiency and socialization in both languages. Finally, I discuss the results in the context of a significant reduction of complexity of the indigenous language and an increased individual, inter- and intra-community variation in Nahuatl and Spanish. I argue the latter reflects varying degrees of bilingualism and proficiency in both languages, modes of their transmission and socialization as well as evolving patterns and spaces of use. My study provides additionally a methodological proposal bridging synchronic and diachronic approaches to the study of long-term history of languages in contact, with special emphasis 16 on varietal distribution, internal variation, levels of language dominance and the continuum of specific features underlying system-altering changes.

Innovations in Brazil in the Pommeranian to-infinitive Accommodation to Portuguese or intradialectal convergence?


While European Portuguese (EP) had a three-fold way to express purpose clauses, Brazilian Portuguese (BP) only has one, but it is unique from the EP options.  In (1a-c), we give the three possibilities in EP. There are two points of variation: variation in the lexicalization of the complementizer üm ‘for’ in C, and lexicalization of the infinitival prefix tau ‘to’ in T. We draw data from the Wenker questionnaire of 1874 in (1a-c) for the locations Schloenwitz, Lankow, and Schlenzig. The fourth logical option, with both C and T empty, does not occur in any location (1d) in EP.
We take the absence of (1d) pattern as an indication that i) the two heads T and C ‘see’ each other at some level of representation, for instance by chain formation (e.g. Pesetsky & Torrego’s T-to-C movement), and ii) that the chain may not remain unexpressed. Curiously, in Brazil neither of these varieties occur. Instead, a new way of expressing the C-T chain has emerged, given under (1e).

(1) e.         du bust nog ni grout naug [taum ain Flasch Wiin ut drinken] (Brazil)

This is an overt way to create the C-T chain and taum lexicalizes the head of this chain.

What is the mechanism of this change? We will develop a parametric model with various choices for the default that implement the two mechanisms and then decide on the basis of external evidence and considerations of plausibility. We assume two (singleton) chains, a for-chain and a to-chain, with values for chain lexicalization and chain reduction. For both models, we assume that lexicalization of a chain is the default, but we make different assumptions on chain reduction: the melting pot mechanism is cast in terms of sensitivity for default values and the accommodation scenario in terms of sensitivity to markedness. (The two models only differ upon multiple or no deletion).

Model 1 - Language internal convergence (“melting pot”) with the assumptions:

       1a. Lexicalization of a chain is the default

       1b. Chain reduction by deletion of the lower copy is the default

Model 2 - Accommodation to the superstrate:

       2a. Lexicalization of a chain is the default

       2b. Chain reduction by deletion of the higher copy is marked

We then get the above table for Model 1 for the 2^3 = 8 logical variants.                          

If we plot the overall markedness, M123 = M1+M2+M3, on the y-axis and represent the x-axis as change over time, the following “energy decay” diagram results.
The melting-pot scenario interprets the change as markedness lowering during the change. A similar table and graph will be presented for the accommodation scenario, and it is then left to the audience to decide which model will be the most plausible for the circumstances at hand.



Pesetsky, D., & Torrego, E. (2001). T-to-C movement: Causes and consequences. Current Studies in Linguistics Series, 36, 355-426.

Language variation and the Wymysiöryś ethnolinguistic identity


Wilamowice is a little town in southern Poland, founded in the Middle Ages by colonists from German(ic)-speaking countries. Wymysiöryś (ISO wym) is a highly endangered language spoken there today by the last few speakers and a growing number of new speakers. After two generations of language decay and attrition, caused by the sociopolitical consequences of the World War II, the language (community) is recently undergoing intensive revitalization and recreation processes.

In the past, Wilamowice formed a part of the so-called Bielsko-Biała linguistic enclave (Bielitz-Bialaer Sprachinsel), which had its roots in the First German Colonization in the 12th/13th century and which, at the climax of its expansion, used to include several villages and towns in the border areas of Silesia and Galicia. For a long time, however, Wilamowice did constitute a peripheral and distinct exclave of the Sprachinsel.

Usually, the Wymysiöryś language is classified as an (East Central) German colonial variety. Both the sociolinguistic research and historical records indicate, however, that at various periods, different opinions on the origin and identity of the community have been considered both by outsiders and by the Wilamowiceans themselves. Such ethnotheories of provenance, including some folk linguistic evidence, referred to Flanders, Holland, Friesland, Low Saxony, even England or Scotland as places of origin of the first settlers. These theories were presented by local men of letters or amateur investigators, such as, for example, Florian Biesik, whose literary output provided numerous facts of both intra-, extra- and folk linguistic character which will be described in this presentation, along with other facts from the sociolinguistic history of Wilamowice.

Considering contact linguistic and dialectological factors, the microlect of Wilamowice has certainly undergone interactions of various type and intensity with Polish (and its varieties) and standard High German. The evidence of such contacts, shift or even hybridization can be found in all subsystems of Wymysiöryś and has contributed to the modern Wilamowicean identity.

The contribution will use some of the paradigms of historical sociolinguistics , combined with the emic /etic approaches of cultural anthropology, to reconstruct the role of language variation factors in shaping the socio- and ethnolinguistic identity of that community and its microlanguage[1].



Andrason, A. (2015) Vilamovicean – A Germanic-Slavic Mixed Language? Studies in Polish Linguistics 10/2: 57–85.

Andrason, A. (2015) Slavic-Germanic hybridization in the Vilamovicean language In Mańczak-Wohlfeld, E & B. Podolak (eds.) Words and dictionaries: 11-27. Kraków.

Wicherkiewicz, T. & J. Olko (2016) Researching, Documenting and Reviving Wymysiöeryś: a Historical Outline. In Olko, J., T. Wicherkiewicz & R. Borges (eds.) Integral Strategies for Language Revitalization: 17:54. University of Warsaw.

[1] Microlanguages are language varieties used by small, geographically or culturally isolated communities of speakers. The geographical isolation may result from peripheral or insular location, while the cultural one from religious, racial, ethnic or social determinants.

The complexification of endangered minority languages in situations of contact: Wymysiörys


Contrary the common view that language contact brings about the simplification of endangered minority languages, this paper demonstrates that it may also be responsible for their complexification. Specifically, it is argued that the complexity of Wymysiöryś has gradually increased due to prolonged interaction with Polish, the dominant language of the area.

Drawing on the ideas of Dahl (2009), to verify the increase of complexity, the complexity of Wymysiöryś is contrasted with the complexities of two diachronic and/or dialectal “control”-systems: Middle High German (MHG) and Modern Standard German (MSG).

The study focuses on the absolute complexity of Wymysiöryś and the two control-systems, where the complexity of language viewed as “an autonomous entity” (Kusters 2008:4; Miestamo 2008). Furthermore, to calculate the absolute complexity of each language, we deploy the concept of eective complexity, which estimates the degree of non-randomness included in a language system and the amount of information required to describe rules governing a given content (Gell-Mann 1995; Miestamo 2008; Sinnemäki 2011).

First, we analyze local complexities (Miestamo 2008) pertaining to diverse domains that are contained in the five major levels of language: phonetics, vocabulary, morphology, morphosyntax, and syntax. These local values are subsequently combined into multi-vectorial global complexity (Deutscher 2009).

The study of local and global complexities demonstrates the following: (a) locally, the complexity of Wymysiöryś is either equal or superior to that of MHG and MSG; (b) globally, the complexity of Wymysiöryś is greater than the overall complexity of MHG and MSG; (c) both locally and globally, the surplus of complexity exhibited by Wymysiöryś can always be attributed to its contact with Polish.

That is, by assimilating various Slavonic properties, and simultaneously maintaining its inherited or internally developed Germanic traits, Wymysiöryś is lexically and grammatically richer than its mother and sister languages. Therefore, the complexity of Wymysiöryś is also likely greater than that of its immediate input–a pre-contact, colonial variety brought to Wilamowice by German(ic) settlers in the 13th century. The subsequent increase in complexity is a direct eect of the Polonization of that “proto”-variety.



Dahl, Östen 2009. Testing the assumption of complexity invariance: The case of Elfdalian and Swedish. In Georey Sampson, David Gil, & Peter Trudgill (eds.), Language Complexity as an Evolving Variable, 50-63. Oxford: OUP.

Deutscher, Guy 2009. ‘Overall complexity’ a wild goose chase? In Georey Sampson, David Gil, and Peter Trudgill (eds.), Language Complexity as an Evolving Variable, 243-251. Oxford: OUP.

Gell-Mann, Murray 1995. What is complexity? Complexity 1(1). 16-19.

Kusters, Wouter 2008. Complexity in linguistic theory, language learning and language change. In Matti Miestamo, Kaius Sinnemäki, & Fred Karlsson (eds.), Language Complexity: Typology, Contact, Change, 3-22. Amsterdam: Benjamins.

Miestamo, Matti. 2008. Grammatical complexity in a cross-linguistic perspective. In Matti Miestamo, Kaius Sinnemäki, & Fred Karlsson (eds.), Language Complexity: Typology, Contact, Change, 23-42. Amsterdam: Benjamins.

Sinnemäki, Kaius. 2011. Language Universals and Linguistic Complexity. PhD dissertation, University of Helsinki.

13:30-15:10 Session 3F: Panel Social Media
Location: New York 3
The normality of online vernacular writing – examples from a Norwegian context


This paper explores the online linguistic practices of young Norwegians from the Trondheim region. Norway is widely known as a comparatively diffuse speech community with a high degree of dialectal variance and, moreover, an extensive tolerance of linguistic variation (cf. Jahr 2014; Røyneland 2009). The idea of Norway as ‘a realm of dialects’ is deeply embedded in the Norwegian population, not least due to a long history of official recognition and protection of linguistic diversity. Even though digitally mediated language use in Norway has been given only limited scientific attention so far, we have clear indications of extensive use of vernacular writing on digital arenas such as SMS, Facebook and messaging platforms. This paper aims to display how young Norwegians conceive of ‘dialect writing’ as a linguistic normality within their everyday literacy. The analyses rest on two data sets with a mixed-methodology design (cf.  Creswell 2009), both comprising authentic written material from social media (viz. Facebook’s Messenger platform and Instagram) and ‘offline’ individual semi-structured interviews with the writers behind the texts in question. Drawing on these ‘blended data’ (cf. Androutsopoulos 2017), I will discuss how we can understand the different linguistic choices in light of the rather special Norwegian ‘sociolinguistic climate’ as acts of identity negotiation. The material is analyzed both within a descriptive variationist framework focusing on the formal aspects of the linguistic practices, and by applying a qualitative approach chiefly inspired by discourse analysis (cf. e.g. Androutsopoulos 2013). The material demonstrates that the informants orient toward different co-occurring norms of language use, where ‘normal Norwegian’ (i.e. the official written norms) and ‘dialect’ are the central options for most informants. My inquiry suggests that the high occurrence of vernacular writing among young Norwegian should be seen against the background of the prevailing cultural model of ‘dialectality’ as a normalcy. This involves a strong general expectation of linguistic conduct that corresponds with the individual speaker’s – and writer’s – (ascribed) local belonging. Any significant mismatch between linguistic practice and affiliation to a place might compromise the individual’s authenticity and thus be socially disapproved. This material shows that such sociolinguistic mechanisms also have effect in written online contexts and thus make ‘dialect writing’ the default choice in many contexts.



Androutsopoulos, J. 2017. Online Data Collection. In: Mallinson, C et al. (eds.), Data Collection in Sociolinguistics. Methods and Applications, 492–520. Routledge.

Androutsopoulos, J. 2013. Participatory Culture and Metalinguistic Discourse: Performing and Negotiating German Dialects on YouTube. In: Tannen, D. & Trester A.M. (eds.), Discourse 2.0: Language and New Media, 47–73. Georgetown University Press.

Creswell, J.W. 2009. Research design: Qualitative, quantitative, and mixed methods approaches. Sage.

Jahr. E.H. 2014. Language Planning as a Sociolinguistic Experiment: The Case of Modern Norwegian. Edinburgh University Press.

Røyneland, U. 2009. Dialects in Norway: catching up with the rest of Europe? International Journal of the Sociology of Language 196-197, 7–30.

Dialect in the media – mediatization and processes of standardization


In this paper, we focus on how young peoples’ engagement in social media practices can inform our understanding of local processes of standardization. Following a growing body of mediatization research (e.g. Agha 2011; Androutsopoulos 2014), we view mediated language use as a central factor in sociolinguistic change in contemporary societies. Analyzing and comparing young peoples’ use of dialect on Facebook across three dialect areas in Denmark (Maegaard et al, forthc.), we argue that language variation in digital language practices is embedded with social meaning and serves interactional functions in similar ways as variation in speech (Androutsopoulos and Ziegler 2004; Stæhr 2016). Our analysis of digital language practices draws on data collected on social media sites, sound recordings made by the participants in various everyday situations and interviews with the participants and their parents and grandparents. Our analytical approach combines quantitative and qualitative analysis of the participants’ digital language practices.

By studying the young peoples’ dialect use across online and offline contexts, we discuss different ways in which online dialect contributes to our understanding of the role of social media in processes of sociolinguistic change (Coupland 2014). We do so by addressing how dialect can be used in similar ways across online and offline everyday contexts and by discussing the ways metalinguistic reflections on dialect spelling contribute to knowledge about norms of everyday dialect use. Finally, we consider how audiovisual mediation of dialect in mass- and social media affects users’ perceptions and evaluations of their local ways of speaking and how this is expressed in meta-pragmatic accounts across generations. Such meta-linguistic reflexivity in the context of everyday mediatization allows us to further discuss the status of the permeating standard language ideology in Denmark. This forms the basis for a more general discussion of the role of online dialect use in processes of standardization. 



Androutsopoulos, Jannis. 2014. Mediatization and sociolinguistic change. Key concepts, Research traditions, open issues. In Androutsopoulos, Jannis (ed.): Mediatization and Sociolinguistic Change. 3–48. Berlin, Boston: de Gruyter (linguae & litterae 36).

Androutsopoulos, Jannis & Evelyn Ziegler. 2004. Exploring language variation on the Internet: Regional speech in a chat community. In Britt-Louise Gunnarsson (ed.) Language Variation in Europe. Papers from ICLaVE 2. 99–111. Uppsala: Uppsala University Press. Coupland, Nikolas. 2014. Sociolinguistic change, vernacularization and broadcast British media. In Androutsopoulos, Jannis (ed.) Mediatization and Sociolinguistic Change. 67–96. Berlin: DeGruyter.

Maegaard, Marie, Malene Monka, Kristine K. Mortensen & Andreas C. Stæhr (eds.). Forthc. Standardization as sociolinguistic change. A transversal study of three traditional dialect areas. Routledge studies in language change. Abingdon: Routledge.   

Stæhr, Andreas. 2016. Normativity as a social resource in social media practices. In Madsen, Lian Malai, Martha Sif Karrebæk & Janus Spindler Møller (eds.): Everyday Languaging. Collaborative Research on the Language Use of Children and Youth. Trends in Applied Linguistics. 71–94. Mouton De Gruyter.

Non-standard spelling in WhatsApp-chats – Local functions and accommodation


The use of non-standard forms has become firmly established in interactive writing, as shown in WhatsApp and earlier in SMS and IRC chats. However, the status of these non-standard forms in the same language also varies considerably depending on the linguistic situation. For example, about 85% of German chats in Switzerland can be classified as dialectal (Ueberwasser/Stark 2017: 112), while the majority of the chats in Germany are oriented towards the standard variety. The use of dialectal forms thus has a fairly different meaning in each case. However, in both countries there are big differences between individual chatters, so that generalizations are to be questioned and scrutinized. The focus of the presentation therefore goes away from general statements on the use of different varieties in chat communication to the investigation of selected chat sequences, which can show functions of standard deviating spelling (and thus also standard oriented spelling). The two corpora "What's up, Switzerland?" and "What's up, Germany?", which each consist of several hundred thousand WhatsApp messages, form the data basis for the study. In a first step, the study is centered on concrete interactions and focuses on the local functionalization of specific spellings and grapho-stilistic variants. On the basis of the results of such local analyses, entire chats can then be examined. On the one hand, this allows to show how the meaning of individual linguistic and grapho-stylistic forms is negotiated in the interaction. On the other hand on can trace how chatters adopt forms of their interlocutors, how they accommodate in the course of a chat, thus expanding or reducing the use of such linguistic and grapho-stylistic forms. When expanding the use of such forms into other chats this, in the long term, can even result in language change (Siebenhaar 2018). By comparing German and Swiss chats, the results can then – very carefully – also be placed in cultural contexts.  


References (Style: Heading 2):

Ueberwasser, Simone & Elisabeth Stark (2017): What’s up, Switzerland? A corpus-based research project in a multilingual country. Linguistik online 84: 105–126.

Siebenhaar, Beat (2018): Sprachgeschichtliche Aspekte der Verwendung von Bildzeichen im Chat. In:Czajkowski, Luise, Sabrina Ulbrich-Bösch & Christina Waldvogel (ed.): Sprachwandel im Deutschen: 307–318. Berlin / Boston: De Gruyter.

Methodological challenges of dialectal data extraction through the net. Two case studies from Greece


The aim of the paper is to present the methodological challenges of two case studies which tried to use the benefits of the internet for crowdsourcing dialectal information and dialectal data.

In recent years, the use of mobile devices’ apps such as Dialäkt Äpp (Kolly et al. 2014), Voice Äpp (Leemann et al. 2015), and English Dialects App (Leemann, Kolly & Britain 2018) in dialectal research has revealed the Internet’s potential as a tool for crowdsourcing a vast amount of dialectological information and data (see also Gonçalves & Sanchez 2014, Leemann et. al. 2016).

In our effort to extend Internet use for dialectological research, we explored two methodological questions: a) How can we classify different geographical dialects crowdsourced from casual recordings on YouTube, and b) can we formulate a digital speech community which will allow us to collect dialectal information and dialectal data?

As to the first question, we ground-sourced YouTube video-files extracting casual dialectal speech. However, we realized that we could not recognize varieties of different dialects other than our own. Nevertheless, when we used a perception test aiming native speakers of particular dialects, we managed not only to dialectally classify the varieties used in the recordings, but also to allocate them to particular communicative situations, age groups and educational levels.

As to the second question, we created a group in FaceBook, asking for the participation of people who were interested in the cultural heritage of her island. The initial reaction of the people was positive; however, a continuous and everyday effort was required so as to keep up the vitality of the group. We realized that even a photograph, or a comment on daily basis was necessary for keeping the group active. Members of the group reacted positively to the researcher’s questions / requests-for-help (40% of the members, 120-130 actual participants). We asked them to participate in google questionnaires/perception tests checking the acceptability of phonological and morphological features of the dialect. The replies revealed interesting patterns of linguistic change, as well as different levels of acceptability in relation to different communicative situations and topics of discussion.


Bibliographic References:

Gonçalves, B. & D. Sanchez. (2014). Crowdsourcing Dialect Characterization through Twitter.  PloS one. 9. 10.1371/journal.pone.0112074.

Kolly, M.-J. & A. Leemann. (2014). Dialäkt Äpp: Communicating dialectology to the public – crowdsourcing dialects from the public. In A. Leemann, M.-J. Kolly, S. Schmid, & V. Dellwo. (Eds.). Trends in Phonetics in German-speaking Europe: 271-285. Bern/Frankfurt: Peter Lang.

Leemann, A. et al. (2015). Voice Äpp: a mobile app for crowdsourcing Swiss German dialect data. InterSpeech 2015: 2804 – 2808.

Leemann, A., M.-J. Kolly, R. Purves, D. Britain, E. Glaser. (2016). Crowdsourcing language change with smartphone applications. PLoS one. doi: 10.1371/journal.pone.0143060

Leemann, A., M.-J. Kolly & D. Britain. (2018). The English Dialects App: The creation of a crowdsourced dialect corpus. Ampersand 5: 1 – 17.

15:10-15:45Coffee Break
15:45-17:25 Session 4A: Perception
Location: Brussels
Accent Bias and Perceptions of Professional Competence in the UK


Unequal outcomes for individuals from less privileged backgrounds in professional hiring have been widely reported in the UK (Ashley et al. 2015). Although accent is one of the most salient signals of such a background in Britain, its role in unequal outcomes remains under-examined. Moreover, there has been almost no large-scale survey of accent attitudes in the UK using audio stimuli (cf. Hiraga 2005 vs. studies of named varieties, e.g. Coupland and Bishop 2007). This paper investigates current public attitudes to accents in Britain, as a first step in a larger project that investigates whether unconscious accent bias plays a role in how job candidates are evaluated.

A nationwide survey using a verbal guise design was conducted with a representative sample of the UK population (n=1014). Participants were asked to evaluate the interview performance of 10 “candidates” for a trainee solicitor position at a corporate law firm in the UK. Stimuli included audio responses to 10 interview questions: 5 requiring legal expertise and 5 focusing on more general skills. The candidates were native speakers of 5 regional English accents (2 speakers per accent):  Received Pronunciation (RP), Estuary English (EE), Multicultural London English (MLE), General Northern English (GNE), and Urban West Yorkshire English (UWYE). These accents were chosen to allow us to examine listener evaluations across a number of fundamental social contrasts conveyed through accent: region (North vs. South), prestige (standard vs. non-standard), localness (local, supralocal, national), age (established vs. newly emergent varieties), and both ethnic- and class-based associations. After hearing each response, participants rated the candidates’ perceived expertise, likeability, professionalism, and hireability. At the end of the survey, participants also provided their opinions about patterns of discrimination in the UK, their tolerance for diversity, and information about their own social, linguistic and demographic backgrounds.

Initial results show a main effect of accent, such that, on average, RP continues to be evaluated most highly, urban London varieties (EE and MLE) receive the lowest ratings, and Northern varieties receive intermediate evaluations. Regardless of accent, speakers received a significant boost in ratings when providing expert legal content in their response, indicating a positive potential for expert knowledge to “override” accent bias. In the talk, we present further fine-grained statistical analyses examining these main effects in relation to additional relevant factors, including listener familiarity with a given accent, listener demographics, and listener attitudes regarding the existence of bias and tolerance of diversity.

Overall, our findings suggest a continuing presence of bias against certain accents in the UK, particularly class- and ethnically-marked urban varieties. The next phase of the wider project from which this talk is drawn will assess the direct effect of these attitudes on hiring within a professional context, with the goal of clarifying the status of accent as a proxy for more general social biases, with negative implications for social mobility.

Correspondences between vowel phoneme boundary locations in production and perception in dialects of North-East England


Research on sociophonetic aspects of the production-perception link indicates that listeners’ placement of phoneme boundaries in two-alternative forced-choice (2AFC) lexical decision tasks depends to some extent upon where the corresponding boundaries fall in their own speech (e.g. Kendall & Fridland 2012, 2017). It may also influence the locations of other phoneme boundaries: recent work on regional US Englishes (Kendall & Fridland 2017) found that the perceived position of the /æ ~ ɑ/ boundary was predictive of where listeners placed the /ɔ ~ ɑ/ boundary in their spoken utterances. In the present study we adopt a similar 2AFC approach to investigate the perception of vowel contrasts in three accents of North-East England (Newcastle, Sunderland, Middlesbrough), using a corpus gathered for the TUULS project (‘The Use and Utility of Localised Speech Forms in Determining Identity: Forensic and Sociophonetic Perspectives’; UK ESRC ES/M010783/1).

The vowels in question are those of the face, goat and nurse sets, which vary markedly within the individual speech communities as well as across the region. In Newcastle and Sunderland, face is regularly produced as [], with a low F1, while in Middlesbrough it tends to be more open, at [ɛː], predicting higher F1. A 6-step face continuum was created in Praat by incrementally adjusting F1 between the [] and [ɛː] poles (382 and 531Hz, respectively) in 24Hz steps, embedded in a sociophonetically neutral /fVz/ frame (as in ‘phase’ ~ ‘fairs’). An analogous [] ~ [ɔː] continuum was generated for goat (/kVt/, as in ‘coat’ ~ ‘court’). In local nurse pronunciations, the key acoustic parameter is F2. Productions in the region range from [ɔː] to [ɛː], with the retracted variant being associated most closely with Newcastle, and the front [ɛː] with Middlesbrough. A third 6-step continuum was created between [ɔː] and [ɛː] by adjusting F2 in 69Hz increments between 1466 and 1813Hz in a /stVz/ (‘stirs’ ~ ‘stairs’) frame.

The test stimuli were then presented to 21 untrained North-Eastern listeners, who judged which word they had heard by choosing between two on-screen options. The inferred locations of the phoneme boundaries for face, goat and nurse were then compared to those in the listeners’ own productions. Our data confirm expected relationships between the boundaries in production and perception. In face the [] ~ [ɛː] boundary differs systematically across the fieldwork sites in line with the distribution of high-F1 face productions, while for nurse we observe a greater acceptance of high-F2 [ɛː]-like tokens as instantiations of ‘stirs’ among Middlesbrough listeners than among participants in the two other groups, in parallel with the typically fronted pronunciation of nurse in spoken Middlesbrough English.



Kendall, T. & Fridland, V. (2012). Variation in perception and production of mid front vowels in the U.S. Southern Vowel Shift. Journal of Phonetics 40: 289–306.

Kendall, T. & Fridland, V. (2017). Regional relationships among the low vowels of U.S. English: Evidence from production and perception. Language Variation and Change 29: 245–271.

15:45-17:25 Session 4B: Computational Sociolinguistics
Location: London
Grammar and Sociopragmatics of Unofficial Dialectal Names in Rural Communicative Networks


Dialectal and unofficial names in which the surname precedes the first name (s Bachmanns Anna, de Schmidte Karl) are used to refer to absent referents in rural speech communities where locals know each other. The presentation introduces the DFG-funded research project “Grammatik und Soziopragmatik inoffizieller Personennamen in Dialekten des Deutschen“, and shows methods and goals as well as first results of the project.

Structural preferences of specific forms can be found in certain dialect areas: We find areas where combinations of surname and first name appear as compounds with or without linking elements (die Müller(s) Anna), in others, they appear as genitive phrases with or without a determiner ((des) Müllers Anna). There are also areal overlaps between variants, and in some local dialects we find several coexisting forms (e.g. phrases and compounds). These cooccurrences can trigger the establishment of diverging contexts of use for different forms. Their usage can be conditioned by sociopragmatic factors, such as the referent’s gender and age, or situations in which speakers intend to disambiguate referents or relations between referents.

Local residence of the referent, ingroup membership of the interlocutors, or formality of speech determine what a person is called, and under which circumstances an unofficial name can change. In the Ripuarian Rheinbach (North-Rhine-Westphalia) for instance, women keep their birth name in front of their first name for their whole life when they have always been living inside the speech community and belong to a resident family of origin. However, when women have moved to Rheinbach due to marriage and their family of origin is not known by other ingroup members, their husband’s surname is placed in front of their first name.



Bach, Adolf (1952): Die Verbindung von Ruf- und Familiennamen in den deutschen, insbesondere den rheinfränkischen Mundarten. In: Meisen, K/F. Steinbach/L. Weisgerber (Hrsg.): Rheinische Vierteljahrsblätter 17 (Mitteilungen des Instituts für geschichtliche Landeskunde der Rheinlande an der Universität Bonn: Ludwig Rohrscheid Verlag, 66-89.

Berchtold, Simone/Dammel, Antje (2014): Kombinatorik von Artikel, Ruf- und Familiennamen in Varietäten des Deutschen. In: Debus, Friedhelm/Rita Heuser/Damaris Nübling (Hrsg.): Linguistik der Familiennamen (Germanistische Linguistik 225-227). Hildesheim, Zürich. New York: Olms und Weidmann, 249-280.

Cornelissen, Georg (2016): "Mit doep- unnd toname?" Personennamen als Teil einer Sprachgeschichte des Dorfes – mit Beispielen vom Niederrhein. In: Roolfs, F.H. (Hg.): Bäuerliche Familiennamen in Westfalen. Münster, 71-81.

Nübling, Damaris (2012): Auf dem Weg zu Nicht-Flektierbaren: Die Deflektion der deutschen Eigennamen diachron und synchron. In: Rothstein, Björn (Hrsg.) Nichtflektierende Wortarten. Berlin/New York (Reihe Linguistik – Impulse & Tendenzen 47). Berlin/New York: De Gruyter, 224-246.

Nübling, Damaris/Mirjam Schmuck (2010): Die Entstehung des s-Plurals bei Eigennamen als Reanalyse vom Kasus- zum Numerusmarker. Evidenzen aus der deutschen und niederländischen Dialektologie. In: Zeitschrift für Dialektologie und Linguistik 77/2, 145-182.

Roolfs, Friedel Helga (2016): Anna Bergmann und Maria Witten. Parentale Femininmovierung von Familiennamen in westfälischen Varietäten. In: Roolfs, Friedel Helga (Hg.): Bäuerliche Familiennamen in Westfalen. Münster, 57-70.

Hella amazingly talented and hella gonna be spreading even more! Regional and linguistic properties of hella in US tweets: the 2018 state-of-the-art


The form hella supposedly originated as a slang term in the California Bay Area in the mid- to late 1980s. (Online) discussions concerning its regional distribution as well as exact functions started then, but the form gained popularity through popular culture only in the late 1990s/earlier 2000s (Le 2013-14). Many linguists likely associate hella with Bucholtz et al.’s 2007 paper: Based on perceptual dialectological differences in California, hella emerged as the most salient marker of “NorCal” at the time (ibid.). Bucholtz (2006/2012:286) and Le (2013-14:45) also mention additional properties (origins in African American English/hip-hop, unmarked and age-graded in Northern California, but marked and “trendy” outside of it), which have received little attention.

Linguistically, hella is extremely versatile: it co-occurs (descending order of frequency, present study) with adjectives, nouns, verbs (including auxiliaries) and adverbs. The 2002 OED entry specifies an adverb (meaning "very") and adjective function ("much, a lot of"). Speculations on hella’s origins mostly assume a univerbation process of (a) hell of (a) (e.g. Bucholtz et al. 2007), while its adoption to other regions seems to follow the intensifier function (Le 2013-14:47).

Surprisingly, no detailed studies exist on the exact (linguistic) usage patterns of hella or the geographical/dialectological extension of its use/spread. Mostly restricted to mentions, hella is discussed concerning theoretical usage patterns without quantitative backing (e.g. Waksler 2000, Le 2013-14). Recently, Eisenstein (2017) mentions that – while having spread all over the US – hella is still proportionately more frequent in Northern California (Twitter data, 2014).

The present paper is based on almost 15,000 geo-coded US tweets containing hella, collected between May and August 2018. The analysis focuses on the variables dialect (region), modification target (focus on N-V-Adj), frequency of the target, and emotional tone of the target (positive-neutral-negative). The data are analyzed with the help of conditional inference trees, logistic regression, keyword analysis, and comparisons of means plus confidence intervals. The results show that

  1. hella is still typical of (Nor)Cal, but also has other epicenters: the geographical spread suggests urban diffusion patterns;
  2. hella combines most frequently with adjectives everywhere, followed by nouns and verbs, but regional differences exist;
  3. vis-à-vis non-Western regions, California is marked by keywords associated with AAE and/or hip-hop;
  4. important statistical results include:
    • the perceived NorCal–SoCal contrast is minimal to non-existent;
    • the most significant regional contrasts exist between the West (based on Labov et al.’s 2006 composite map) and the remainder of the US, not between California and other regions, suggesting recent geographical diffusion;
    • hella+N is typical of the non-West, while hella+V occurs more typically in the West, indicating advanced (de)grammaticalization of the form there;
    • frequency effects play out differently in different contexts and regions; thus, in NorCal, frequency behaves linearly, with higher-frequency adjectives more commonly employing hella, while the reverse is true for the non-Western regions;
    • semantically, negative meanings (still?) dominate for all POS in NorCal, while the “new” hella users outside the Western US overall prefer neutral or even positive collocates.



Bucholtz, M. (2012[2006]). Word Up: Social Meanings of Slang in California Youth Culture. In A Cultural Approach to Interpersonal Communication: Essential Readings, Jane Goodman and Leila Monaghan (eds.), 243-267. Malden: Blackwell.

Bucholtz, M. et al. (2007). Hella Nor Cal or Totally So Cal? The Perceptual Dialectology of California. Journal of English Linguistics 35(4): 325-352.

Eisenstein, J. (2017). Identifying regional dialects in online social media. In Charles Boberg, John Nerbonne and Dominic Watt (eds.), Handbook of Dialectology, 368-383. Oxford: Wiley-Blackwell.

Labov, W., S. Ash and C. Boberg. (2006). The Atlas of North American English: Phonetics, Phonology, and Sound Change. Berlin/New York: Mouton de Gruyter.

Le, P. (2013-14). Hella Controversial: A summary of both the grammatical and societal role of the slang term hella. Prized Writing 2013-14: 41-49. http://prizedwriting.ucdavis.edu/2013-2014. Accessed 2018-05-24.

Waksler, R. (2002). A HELLA New Specifier. International Linguistics Association, 45th Annual Meeting Georgetown University, April 7-8, 2000. https://linguistics.ucsc.edu/research/publications/Hankamer%20Webfest/waksler.html. Accessed 2018-06-07.

Investigating morphological variation on Twitter: Feminine possessive complements of locative adverbial constructions in Peninsular Spanish varieties


Scholars have recently begun to study morphosyntactic variation in Spanish locative adverbial phrases between prepositional complements (i.e. ADVERB + PREPOSITION de + PERSONAL PRONOUN, e.g. delante de mí) and innovative possessive complements (i.e. ADVERB + POSSESSIVE PRONOUN, e.g. delante mío) (cf. Marttinen Larsson & Bouzouita, 2018, inter alia). However, further variation in the suffixes of the possessive pronouns in locative phrases (e.g. delante mío vs delante mía) remains quite understudied. From a normative point of view, it is argued that the use of the -a variant constitutes a less frequent and non-prestigious variant than its masculine -o counterpart (RAE/ASALE, 2009: §18.4o). It has also been attested that the possessive -a variant is originated in and limited to Peninsular Spanish, with a particularly frequent use in Andalusian Spanish (cf. Marttinen Larsson & Álvarez López, 2017; Salgado & Bouzouita, 2017). Scholars have proposed a variety of hypotheses regarding the functioning of this case of morphological alternation. These hypotheses range from the morphological variation being conditioned by agreement with the gender of the alluded referent (cf. Zamora Vicente, 1967: 433) to analogy with the final vowel of the preceding adverb (cf. Alcina Franch & Blecua, 1975: 619-620; inter alia) or linguistic contact with neighboring languages such as Galician (cf. Marttinen Larsson & Álvarez López, 2017: 88-89). However, the available data has been scarce so far, thus not allowing for a quantitative analysis of the innovative feminine variant’s use in different European varieties of Spanish.

In the present study, I examine the possible effect of external factors (e.g. diatopy, etc.) and internal factors (type of adverb, grammatical person, grammatical number, etc.) on this case of variation in Andalusian Spanish. I use Twitter for the data collection. This data gathering method is, essentially, at its very beginning in linguistics, especially within the area of Hispanic linguistics. Through the R package rtweet (Kearney, 2018), the searches are centered on various regions in Andalusia, allowing for a deeper understanding of the use of the -a variant from a synchronic perspective.



Alcina Franch, J. & J. Blecua (1975). Gramática Española. Barcelona: Ariel.

Kearney, M. (2018). rtweet: Collecting Twitter Data. R package version 0.6.7. https://cran.r-project.org/package=rtweet

Marttinen Larsson, M. & L. Álvarez López (2017). ‘Delante suyo’ vs ‘Delante de él’: el uso de las locuciones adverbiales locativas desde una perspectiva diacrónica y diatópica. Signo y Seña, 31(1). 85-104.

Marttinen Larsson, M. & M. Bouzouita (2018). Encima de mí vs. encima mío: un análisis variacionista de las construcciones adverbiales locativas con complementos preposicionales y posesivos en Twitter. Moderna språk 112(1). 1-39.

Real Academia Española/Asociación de Academias de la Lengua Española (2009). Nueva gramática de la lengua española, vol. I-II. Madrid: Santillana.

Salgado, H. & M. Bouzouita (2017). El uso de las construcciones adverbiales locativas con pronombre posesivo en el español peninsular: un primer acercamiento diatópico. Zeitschrift für romanische Philologie, 133(3). 766-794.

Zamora Vicente, A. (1967). Dialectología española. Madrid: Gredos.

needs+PAST PARTICIPLE in regional Englishes on Twitter


The use of needs with a past participle (e.g., ‘The car needs washed’) has been identified as a diagnostic grammatical feature of the US Midland dialect region (e.g., Murray & Simon 2006). Trudgill (1983:16) has also described needs+PAST as a feature of Englishes in Ireland, Scotland, and northern England, and the presence of the construction in the US Midland is typically attributed Ulster Scots settlement in United States (e.g., Montgomery 1991:183).

Studying the feature in the United States and United Kingdom has proven difficult, however. Needs+PAST occurs infrequently in speech, and so cannot be studied with traditional variationist methods. Most studies of needs+PAST have therefore relied on grammaticality judgments from survey respondents (e.g., Murray et al. 1996). But Labov et al. (2006:296) caution that people are generally unable to assess their usage of needs+PAST, so these grammaticality judgments are unreliable.

The tremendous volume of speech-like text on Twitter offers a solution to the challenge of collecting production data for low-frequency variables like needs+PAST. Expanding on methods demonstrated by Doyle (2014), this study reports from a corpus of 3.6 million tweets containing a form of need and collected from 50 cities in the United States, United Kingdom, and elsewhere. A subset of 3291 tweets are identified as containing needs+PAST, and these are quantified and mapped to create a global map of the distribution of needs+PAST in Englishes and the densities of needs+PAST relative to needs+TO.

Results confirm needs+PAST as a productive feature of Scotland, Belfast, Newcastle, and the US Midland, and support claims that the construction spread via immigration. In doing so, it validates studies based on grammaticality judgments, while also demonstrating new techniques for studying productions of low-frequency linguistic variables. Results also provide quantitative evidence of the extent to which a settler variety of English may leave an imprint of itself over several centuries, and of the durability of regional dialect boundaries over long periods of time. The study concludes by suggesting new directions for examining needs+PAST based on qualitative evaluations of production data.



Doyle, G. (2014). Mapping dialectal variation by querying social media. In S. Wintner, S. Goldwater, & S. Riezler (Eds.), Proceedings of the 14th Conference of the European Chapter of the Association for Computational Linguistics: 98-106. Gothenburg, Sweden: Association for Computational Linguistics. Retrieved from http://aclweb.org/anthology/E/E14/E14-1011.pdf

Labov, W., Ash, S., & Boberg, C. (2006). The atlas of North American English: Phonetics, phonology and sound change. Berlin: De Gruyter.

Montgomery, M. (1991). The roots of Appalachian English: Scotch-Irish or British Southern? Journal of the Appalachian Studies Association, 3, 177-191.

Murray, T. E., Frazer, T.C., & Simon, B. L. (1996). Need+past participle in American English. American Speech, 71, 255-271.

Murray, T. E., & Simon, B. L. (2006). What is dialect?: Revisiting the Midland. In T. E. Murray & B. L. Simon (Eds.), Language variation and change in the American Midland: A new look at “Heartland” English: 1-30. Amsterdam: John Benjamins.

Trudgill, P. (1983). On dialect: Social and geographical perspectives. Oxford: Blackwell.

15:45-17:25 Session 4C: Panel Coherence
Location: Madrid
Two processes of word-final n-deletion in Central Franconian dialects, their distribution, their conditioning and what keeps them together, underground and at the surface


With the exception of the varieties spoken in the southwest and the northeast, all varieties of spoken Dutch have a late process of the deletion of post-schwa word-final /n/ in suffixes (mak-en, to make, het lev-en, the life etc). The process is near categorical and almost entirely insensitive to stylistic factors. Neighbouring Central Franconian dialects of German equally delete post-schwa word-final /n/, both in suffixes and in stems – albeit allophonically.

The Central Franconian dialect area covers parts of Germany, France, Luxembourg, Belgium and the Netherlands. Today, the dialects are part of a range of different constellations of standard, dialect and intermediate  varieties. The northern branch of the group is known as Ripuarian dialects, the southern branch as Moselle Franconian, A phenomenon most Central Franconian share is the deletion of word-final /n/ after full vowels. Following long vowels (e.g. the dialect equivalents of een, ‘one’, steen, ‘stone’, been, ‘leg’) deletion is either excluded or lexicalised; after short vowels (e.g. the dialect equivalents of ben, ‘[I] am’, van, ‘of’ preposition, deletion is either allophonic (Moselle Franconian) or shows quantitative variation (Ripuarian). The  conditions of word-final /n/ after full vowels invariably  concern  phonological,  prosodic,  morphological  and  sociolinguistic parameters. The deletion  of word-final /n/ after full vowels does not occur in any other dialects of Dutch or German. 

In this talk we will zoom in on the differences and similarities in the distribution of both processes of word-final /n/ deletion in a carefully balanced sample of Central Franconian dialects. From in-depth analyses the processes appear to be connected by two prosodic blocking conditions. This fact, along with a set of properties that are visible at the surface, brings about coherence between both deletion processes and thus, more generally, between the relevant dialects, which are more and more subject to divergent forces brought about by the various standard language regimes they are subject to.   

Multiple Factorial Analysis of hierarchical structures among the variants of four lexical sets in the DECTE corpus.


The present study investigates linguistic variation in Tyneside English (TE) using multiple factorial analysis (MFA) to examine the vocalic system of speakers in the Diachronic Electronic Corpus of Tyneside English (DECTE, Corrigan et al. 2012). At present, MFA is more commonly associated with research in the social sciences and sensory data analysis (Husson et al. 2011), and has hitherto never been used in language variation. Akin to Principal Component Analysis (PCA), MFA has the advantage of reducing the complex dimensionality of language variation data and of showing which linguistic features tend to co-occur by providing clear visual representations. Combined with clustering analysis, MFA provides key insights into the dynamics of inter- and intra-group variation and helps spot outliers. MFA is a powerful tool for variationists for two main reasons. Firstly, it is possible to include both nominal and numeric data into the same analysis, contrarily to PCA. Secondly, it takes into account hierarchical structures in the data by allowing groupings of variables in the analysis. Even if groups of variables contain unequal numbers of variables, the weight of each group is balanced out, whilst in a traditional analysis, “all variables would be dominated by the group with the strongest structure” (Abdi & Valentin 2007).

Our study investigates the co-occurrence of the vocalic variants FACE, GOAT, PRICE and MOUTH in the TLS sub-corpus of DECTE. Initial multivariate analyses were carried out in the 1980s (Jones-Sargent 1983) but not all variables could be fed into the model simultaneously. A more recent study was made (Moisl & Maguire 2008, Corrigan et al. 2014) but the hierarchical structure of the data was not taken into account. The data is composed of frequencies within 63 groups of transcribed phonetic features (n=566, consonants & vowels) for 44 speakers from urban Tyneside. Each vowel variant was considered a variable. The data variables were normalised by computing within group frequencies (all frequencies within a group add up to 1 for each speaker). The MFA results suggest that FACE is a major phonetic determinant of TE. A clustering analysis based on MFA results revealed a major gender and class gap with sub-groups characterising varying degrees of accentedness. FACE and GOAT variants share similar patterns of variation and work in lockstep in a similar way to another DECTE sub-corpus (Watt 1998), while the realisation of the PRICE and MOUTH vowels show more independent patterns.



Abdi, H., & Valentin, D. (2007). Multiple factor analysis. In Salkind, N. J. (ed.): Encyclopedia of measurement and statistics: 657-663. Thousand Oaks (CA): Sage.

Corrigan, K. P., Buchstaller, I., Mearns, A. & Moisl, H. (2012). The Diachronic Electronic Corpus of Tyneside English (DECTE). Newcastle University. http://research.ncl.ac.uk/decte (Accessed 10 January 2019).

Corrigan, K. P., Mearns, A., & Moisl, H. L. (2014). Feature-based versus aggregate analyses of the DECTE corpus: Phonological and morphological variability in Tyneside English. In B. Szmrecsanyi & B. Wälchli (Eds.), Aggregating dialectology, typology, and register analysisLinguistic variation in text and speech (pp. 113-149). Berlin: De Gruyter.

Husson, F. (2011) Exploratory multivariate analysis by example using R. Boca Raton: CRC Press.

Jones-Sargent, V. (1983) Tyne bytes: A computerised sociolinguistic study of Tyneside. Frankfurt am Main: Peter Lang.

Moisl, H. L., & Maguire, W. N. (2008). Identifying the main determinants of phonetic variation in the Newcastle Electronic Corpus of Tyneside English. Journal of Quantitative Linguistics, 15(1), 46-69. 

Watt, D. (1998) Variation and change in the vowel system of Tyneside English. (Unpublished doctoral dissertation), University of Newcastle, Newcastle Upon Tyne.

15:45-17:25 Session 4D: Panel Dialectometry
Location: Moscow
Latvian dialects and their classification


The Latvian language belongs to the Indo-European family and is one of the two surviving languages of the Baltic branch. Latvian has still retained the historical regional dialects, which are subdivided in smaller units, or subdialects, whose boundaries in Latvia coincide with the boundaries of small rural districts established in 1939. The territories of the rural districts historically overlap with former estates and, in many cases, with parishes. Contemporary Latvian has approx. 500 small regional language varieties, or subdialects, which traditionally are grouped into 3 dialects. The Central dialect is spoken in the central part of Latvia. This dialect gave rise to the standard variety of Latvian. Livonian dialect is spoken in the north-west of Vidzeme and northern Kurzeme. The dialect spoken in the eastern part of Latvia is called High Latvian dialect. It has many distinctive features related to sound changes, morphological peculiarities and the lexicon.

Yet the boundaries of the Latvian subdialects and dialects were not clearly demarcated. In 1933 the Baltic linguist Janis Endzelins wrote: “Only when the dialects have been thoroughly investigated, isoglosses (boundaries) of specific features of subdialects could and should be identified and subdialects grouped. The isoglosses should be concerned not only with grammatical but also with lexical features because words in subdialects might be different...” Thus, in J. Endzelīns’ view, a traditional basis of classification of Latvian dialects involved the totality of phonetic, morphological and lexical isoglosses.  The sub-dialects were traditionally described, analysing their morphological, phonetic and lexical qualities until the end of the 20th century.

In early 21st century a project, dedicated to the study of Latvian dialects in a sociolinguistic aspect, was carried out. Special attention was payed to the link between the regional language variant and the local cultural environment, the perception of the sub-dialect as a marker of identity, the vitality of sub-dialects in areas with a functioning school, culture institutions, church, etc.

The language of communication between three different generations, the age-graded linguistic change, the differences in the use of sub-dialect in central and peripheral parts of respective areas, as far as  determining the dynamics of isoglosses of dialectal features were analized.

The results let to draw a conclusion, that nowadays the isoglosses of particular dialectal features are difficult to distinguish due to active migration of the people. The research results of the recent years show that, from one hand, many dialectal features are suffering attrition. From the other hand,  self-preservation tendency of the dialects is observed.

It is difficult to foresee the vitality of sub-dialects in the future, because of attrition of dialectal features and change of their boundaries, e.g., the boundary of the High Latvian dialect has moved  to the East, and its territory is decreasing.

The results of such investigation of the regional dialects  can be used in the regional studies to predict the possible  ways in developing of regional linguistic variety.  

On the classification of Portuguese dialects


As in many other languages areas, the division and classification of the regional varieties spoken in the Portuguese territory has evolved over the last century in tandem with the availability of new resources and new methods (see Ferreira 1994). Also, it has integrated the advances of descriptive linguistics and, to some extent, new theoretical concerns.

While the first dialectal classifications, from the late 19th and the early 20th centuries (Leite de Vasconcelos [1893]1897 and subsequent work), mainly drew dialectological borders on the basis of the geography of administrative areas, the first systematic inquiries for a linguistic atlas in Portugal (as part of the network for the Linguistic Atlas of the Iberian Peninsula - ALPI) made it possible to rely on isoglosses for the identification and delimitation of the Portuguese dialects (Cintra 1971). An important consequence of this empirical shift is evident in this dialectal classification, thus reduced to a bipartite hierarchical division drawn from a selection of isophones. More recently, data from the national geolinguistic project, partially published as Atlas Linguístico-Etnográfico dos Açores - ALEAç, have allowed the extension of this traditional approach to the classification of the Azorean dialects (Segura 2006). Meanwhile, since Saramago 1986 and Vitorino 1988, the availability of other geolinguistic materials have fed dialectometrical approaches to (aspects of) the classification of Portuguese dialects, and it is expected that more comprehensive dialectometric analyses can still be further developed. Also, concomitantly to the compilation of a dialect corpus for the study of syntax (CORDIAL-SIN), the classification of Portuguese dialects has also been considered from the angle of syntactic variation.

This paper is focused on the main linguistic features that can define dialectal areas in the Portuguese territory, taking into account traditional works of Portuguese dialectology and the recent developments in dialectometrical analyses and research on dialect syntax. Special attention will be given to the interplay of different structural elements (phonological, morphological, syntactic and lexical ones) and their contribution to the identification of dialectal borders. Finally, as previous works have emphasized, it will be shown that linguistic variation over the geographical space often fits or relates to other geographical patterns of cultural and historical differentiation.



ALEAç–Atlas Linguístico-Etnográfico dos Açores (coord. J. Saramago), CLUL-CCA/SEC-Açores. http://www.culturacores.azores.gov.pt/alea/

ALPI–Atlas Lingüístico de la Península Ibérica (coord. T. Navarro Tomás), Madrid: CSIC. (see also: www.alpi.csic.es)

Ferreira, M. Barros (1994): Retrospectiva da Dialectologia Portuguesa. Revista Internacional de Língua Portuguesa 12: 108-118.

Cintra, L. F. Lindley  (1971): Nova proposta de classificação dos dialectos portugueses. Boletim de Filologia XXII: 81-116.

Leite de Vasconcelos, J. (1893) Carta dialectológica do Continente Português. Lisboa.

Saramago, J. (1986): Différenciation lexicale (ou essai dialectométrique aplique aux matériaux portugais de l’ALE. Géolinguistique II: 1-31.

Segura, L. (2006) Dialetos Açorianos. Contributos para a sua classificação. In Bernardo, C. & H. Montenegro (eds.): I Encontro de Estudos Dialetológicos – Actas (2003), 325-344. Ponta Delgada: ICPD.

Vitorino, G. (1988) Atlas Linguístico do Litoral Português (ALLP) - Fauna e Flora. Lisboa: CLUL/INIC.

Review of the dialectal classifications in Europe


This topic has had my interest for a very long time, being originally trained as a cartographer. I considered various approaches to the classification of language varieties (also in work of my students, who wrote MA- and BA-papers about it). I experimented with various statistical clustering techniques but also subjective judgements of speakers. An addition that I tried to make to the discussion is the analysis of historical dialect boundaries. One may ask: what is the point in painstakingly trying to define the location of a dialect boundary, once we realize that they are moving from generation to generation anyway? When dialect boundaries are mobile, any dialect classification or division will be no more than the observation of a temporal configuration.

Another phenomenon that stood out to me is that mapping different linguistic domains can lead to quite different distributions, especially when trying to dissect a language continuum. At the same time: clear cut divisions are hardly ever missed, irrespective of the method and linguistic domain chosen, albeit that the hierarchical organization of dialects/varieties may differ from the traditional ones. To make the last point clear: in a wider perspective of Frisian, Dutch and German dialects, Frisian varieties from West, East and North Friesland are unlikely to be primarily associated with each other (the historical division) but will locally always be differentiated from their neighbourhood.

An interesting point will come from the study of secondary links between varieties: the primary groupings mostly reflect the axioma that linguistic vicinity reflects geographical vicinity, but statistically secondary links, or results ‘cleaned’ for vicinity effects, may reveal other and sometimes older relations between languages as in Gooskens and Heeringa (2004), who found that despite the fact that Frisian is much more similar to Dutch than to English, and English is very different from all other Germanic languages, Frisian it is still the closest, or – perhaps one should say –  the least distant relative of English on the continent.



Gooskens, C. & Heeringa, W. (2004). The Position of Frisian in the Germanic Language Area. In, Dicky Gilbers & Maartje Schreuder (ed.): On the Boundaries of Phonology and Phonetics: 61–87. Groningen: University of Groningen.

Versloot, A. P. (2002): Cor en Geer Hoppenbrouwers, De indeling van de Nederlandse dialecten. Us Wurk 51, nr. 3–4: 159–62.

Versloot, A. P. & Adamczyk, E. (2017). Geography and Dialects in Old Saxon: River-basin communication networks and the distributional patterns of North Sea Germanic features in Old Saxon. In Hines, J. & IJssenagger, N. (ed.): Frisians and their North Sea Neighbours, from the fifth century to the viking age: 125–48. Woodbridge: Boydell and Brewer.

Versloot, A. P. (2018). Historical Dialectology – West Frisian in Seven Centuries. In Stanley, B. & Kehrein, R. (eds.): Handbook of the Changing World Language Map. Springer, 2018/forthcoming.

15:45-17:25 Session 4E: Panel Minority
Location: Paris
The socio-linguistic history of Zeelandic-Flemish in Brazil: A case of language decay


Between 1858 and 1862 approximately 330 Dutch settlers from the southern province of Zeeland migrated to Espírito Santo, Brazil. The Zeelandic-Flemish speakers have had diffculty integrating into Brazilian society ever since. More than 150 years later, through field research and numerous interviews in several municipalities of Espírito Santo, we discovered that there are less than 20 speakers remaining. Their original native language, the Zeelandic-Flemish dialect, is critically endangered (UNESCO, 2003). Only a few speakers in the grandparental generation know the language, but do not use it anymore. Written materials do not exist and the language is undocumented (Schaffel-Bremenkamp, 2010).

Several factors contribute to language replacement and language decay, including the small number of Dutch settlers, the isolation of the Dutch community, intermarriage (mostly with Pomeranians), internal migration, a forced change of religion from Calvinism to Lutheranism, and formal education. Zeelandic-Flemish was threatened both by Brazilian Portuguese, and Brazilian Pomeranian, a lingua franca among the Germanic Protestant community in the region. The Pomeranians came to Brazil at the same time for similar reasons and went to the same imperial colony. Zeelandic and Pomeranian immigrants were in regular contact and Zeelandic was gradually replaced by Pomeranian in home conversation.

The case of Brazilian Zeelandic in Espírito Santo is a typical situation of gradual language death (cf. Dorian, 1981; Dressler, 1972; Holloway, 1997), in which all of its speakers show signs of language attrition. Long periods of disuse of a language almost always lead to reduction on all linguistic levels. This structural reduction of the dying language is probably brought about by multiple causes, including transfer from the dominant language and internally-motivated change.

In this presentation we will provide socio-linguistic descriptions of the language acquisition and language replacement processes and provide a preliminary look at how Brazilian Zeelandic exhibits structural reduction (relative pronoun neutralization, do-support, topic drop, complementizer fusion, and loss of diminutives) and to what extent transfer from Pomeranian (and Portuguese) is involved (intra- and interspeaker allophony, lexical borrowing and calquing) (Schaffel-Bremenkamp et al., 2017).



Dorian, Nancy (1981) Language death: The life cycle of a Scottish Gaelic dialect. Philadelphia: UPenn Press.

Dressler, Wolfgang (1972): On the phonology of language death. Chicago Linguistic Society 12: 448-457.

Holloway, Charles (1997) Dialect death. The case of Brule Spanish. (Studies in bilingualism 13). Amsterdam/Philadelphia: Benjamins.

Schaffel-Bremenkamp, Elizana. (2010) Análise sociolinguística do desaparecimento da língua holandesa no Espírito Santo. Trabalho de Conclusão de Curso Departamento de Línguas e Letras, Universidade Federal do Espírito Santo, Vitória.

Schaffel-Bremenkamp, Elizana. et al. (2017): Zeeuws-Flemish in Brazil. Gragoatá: Revista dos programas de pós-graduação dos institutos de letras da UFF, Rio de Janeiro, v. 22, n. 42: 435-472.

UNESCO. (2003) Language Vitality and Endangerment. Available at: http:// www.unesco.org/culture/ich/doc/src/00120-EN.pdf. Accessed on 21/09/2010.

Language use and variation within the Greek linguistic minorities of Southern Italy


Greko and Griko are endangered Italo-Greek languages spoken in the south of Italy. In recent decades, both have undergone a sharp decrease in the number of speakers due to growing influence from two local Romance varieties – southern Calabrese and Salentino –  and, more recently, from regional Italian. In this contribution, we report on preliminary outcomes from our SMILE (Sustaining Minoritized languages in Europe) project. Specifically, we will focus on two different cases of language variation. The first one refers to temporary changes in the language depending on the addressee’s competence, age, and community membership. We shall highlight how Griko/Greko speakers may adapt their linguistic production according to these three major factors; for instance, they may resort to Romance and/or Italian borrowings over Griko/Greko equivalent ones assuming this way to facilitate comprehension by the addressee; or, to the same end, speakers with knowledge of Modern Greek may resort to it. The second case of language variation and change concerns the long-lasting contact with the local Romance varieties and to a certain extent with Italian, as well as the more recent contact with Modern Greek facilitated by tourism, twinnings, and the availability of Modern Greek courses in both regions. This intense contact with multiple languages affects the lexis and the morpho-syntax of Griko and Greko.

This investigation will therefore shed light not only on language changes occurring as a result of contact with different languages and of speakers’ multiple linguistic resources (Italian, Romance dialect, Griko/Greko, Modern Greek), but also on how multilingualism continues to act locally as a resource, which at the same time produces and is produced by the social dynamics at play within the communities.

15:45-17:25 Session 4F: Panel Social Media
Location: New York 3
Evolution of Frisian bilingual teenagers' language use on social media


Since 2013, the use of Frisian on social media has been studied (Jongbloed-Faber 2014; Jongbloed-Faber et al. 2017; 2016; McMonagle et al. 2018). The studies show that despite a low writing proficiency and a linguistically diverse audience, the majority of Frisian speakers use Frisian on social media to some extent. A 2013-2014 study (n=2,267) revealed that on WhatsApp, Frisian is used the most: 55% of the L1 teenagers use it often or all the time. Teenagers’ peer group, language attitudes, and one’s writing proficiency are strong factors (R2=56%) that explain the use of Frisian on social media (Jongbloed-Faber et al. 2016). As the use of social media is volatile, longitudinal studies are needed. Therefore, new research among Frisian teenagers will be conducted early 2019.

The research questions that will be addressed in this paper are: How, when and why are Frisian varieties used on social media by teenagers and how has the use evolved between 2013-2014 and 2019? While WhatsApp, where Frisian varieties were used the most, has remained popular among teenagers, the digital natives’ shift from Facebook and/or Twitter to the image-focused Instagram and/or Snapchat may have repercussions on the use of linguistic resources. On the one hand, many posts may simply lack linguistic resources. On the other hand, as image will be the most important semiotic resource in the post, it may be perceived to be less problematic when the text in a post is written in a variety that cannot be understood by one’s complete audience. The question therefore is how local language varieties are used on these image-focused social media. To explain the variation in language use, I will use the Audience Design framework (Androutsopoulos 2014; Bell 1984) as a starting point.

To answer the research questions, a questionnaire will be distributed among teenagers between 14-18 years following secondary general and vocational education in Fryslân. The quantitative analysis will be enriched by a qualitative analysis of teenagers’ practices on social media.



Androutsopoulos. J. (2014). Languaging when contexts collapse: Audience design in social networking. Discourse, Context and Media 4-5 : 62-73.

Bell. A. (1984). Language style as audience design. Language in Society, 13(2): 145-204.

Jongbloed-Faber. L. (2014b). Social media: a treasure trove for minority language research. In K. Woodfield (ed.), Social Media in Social Research: Blogs on Blurring the Boundaries: 189-194. London: Natcen Social Research.

Jongbloed-Faber, L., Van de Velde, H., Van der Meer, C. & Klinkenberg, E.L. (2016). Language Use of Frisian Bilingual Teenagers on Social Media. Treballs de Sociolingüística Catalana 26: 27-54. DOI: http://dx.doi.org/10.2436/20.2504.01.107

Jongbloed-Faber, L., van Loo, J., & Cornips, L. (2017). Regional Languages on Twitter: A comparative Study between Frisian and Limburgish. Dutch Journal of Applied Linguistics 6(2): 174–196. DOI: https://doi.org/10.1075/dujal.16017.jon

McMonagle, S., Cunliffe, D., Jongbloed-Faber, L., & Jarvis, P. (2018). What can Hashtags tell us about Minority Languages on Twitter? A Comparison of #cymraeg, #frysk and #gaeilge. Journal of Multilingual and Multicultural Development. DOI: https://doi.org/10.1080/01434632.2018.1465429

Dialect use in German WhatsApp chats: from indexing regional identity to (self)stigmatization


Previous research on the use of German dialects (in a broad sense) in digitally mediated communication provides empirical support for two main claims (cf. Androutsopoulos 2010, Androutsopoulos/Ziegler forthc.): First it suggests a range of usage patterns that vary by dialect region and digital context but can nonetheless be typified in a three-part classification in terms of the relative extent of regionally marked language use and its communicative functions online. More specifically, regional language online can be used as the ‘base language’ of a digital speech event, or as a ‘code’ that is juxtaposed to other codes in the speech event, or as an emblem (Agha 2007), whereby single dialect features assume third-order indexical meaning (Johnstone et al. 2006) relative to a user group or digital space. The second claim is that the extent of written (i.e. typed) representation of regional language in German online data is often limited to a few salient features for a given dialect or regional variety and at the same time correlates to the degree of publicness/privacy of an online speech event.

Against this backdrop, this paper examines to what extent these patterns of regional usage hold true for recently collected data that originate in chat groups enabled by WhatsApp. The analysis is based on a corpus of 170 chats with 2,000 postings that can be searched via the Mobile Communication Database (MoCoDa), a new platform for crowdsourced data collection (cf. mocoda.spracheinteraktion.de). We focus on data from users who live in the Ruhr Area, a metropolitan region in the north-west of Germany with approximately 5,3 million inhabitants. Its regional language, Ruhr German, is a socially stigmatized variety that has hardly been researched in modern German dialectology to date. Our analysis will examine whether certain features of Ruhr German can take on regional and social indexical meanings in these online interactions, which linguistic features these are, and how exactly they are being deployed.



Agha, Asif (2007) Language and Social Relations. Cambridge: Cambridge University Press.

Androutsopoulos, Jannis K. (2010) The study of language and space in media discourse. In: Peter Auer & Jürgen Erich Schmidt (eds.) Language and Space: An International Handbook of Linguistic Variation. Vol. I: Theory and Methods, 740758. Berlin/New York: de Gruyter (HSK series).

Androutsopoulos, Jannis K. & Evelyn Ziegler (forthc.) Medien und areale Sprachvariation des Deutschen. In: Jürgen Erich Schmidt & Joachim Herrgen (eds.): Sprache und Raum – Deutsch. Berlin/New York: de Gruyter (HSK series).

Johnstone, B., Andrus, J. & Danielse, A. E. (2006) Mobility, indexicality and the enregisterment of ‘Pittsburghese’. Journal of English Linguistics 34, 77-104.

Beißwenger, M. / M. Fladrich / W. Imo / E. Ziegler (2018) News from the MoCoDa2 corpus: A design and web-based editing environment for collecting and refining data from private CMC interactions. In: R. Vandekerckhove et al. (eds.): Proceedings of CMCCORPORA 2018, 10-14. Antwerp: University of Antwerp. https://www.uantwerpen.be/images/uantwerpen/container49896/files/proceedings_CMCcorpora2018.pdf

The use of Limburgish (the Netherlands) on social media, and its role in local identity construction


This presentation will address two research questions: how do actors position themselves, and which oppositions do they create through language choice and mixing in relation to localities in the Limburg community and larger Dutch society?, and how is place-making of Limburg and its localities on social media achieved through languagecultural practices?

In order to address the first question, results will be presented generated by an automatic language identification tool that was created to classify tweets in the Netherlands in Limburgish, Frisian, Dutch and English among 10,434 tweets (Nguyen et al. 2015). This tool made it possible to identify why Limburgish as a regional language is selected on Twitter, and how Limburgish is recognizable (Nguyen & Cornips 2016). The Nguyen et al. 2015 study reveals that Twitterers accommodate in language choice and mixing to their audiences, that is, users tweet in Limburgish when the addressed Twitterers are known to often use Limburgish as well, and when the tweet itself was written in Limburgish. Moreover, users may start in Dutch to maximize their audience but switch to Limburgish during the conversation (Jongbloed et al. 2017).

Regarding the second research question, the use of Limburgish on social media is crucial in the (re)production of local identities. A good example was the Facebook website Nine Gag op z’n Limburgs ‘Nine Gag going Limburgs’. Social media are commonly considered as a global infrastructure only which allow actors to ‘escape’ their local context through language choice (Leppänen 2012.) and language mixing (for instance with English as a global language) and linguistic hybridity (Androutsopolous 2006, Moll 2014). However, the Facebook website Nine Gag op z’n Limburgs interact at local scale only (Leppänen et. al. 2009:1081). I will show how actors on this website construct Limburg versus other ‘places’ and ‘people’ through a multimodality of images and texts.



Androutsopolous, Jannis. 2006. Multilingualism, diaspora, and the Internet: Codes and identities on German-based diaspora websites. Journal of Sociolinguistics: 520-547.

Jongbloed-Faber, Lysbeth, Jolie de Loo and Leonie Cornips. 2017. Regional languages on Twitter: a comparative study between Frisian and Limburgish. Dutch Journal of Applied Linguistics 6 (2):174–196

Leppänen, Sirpa. 2012. Linguistic and Generic Hybridity in Web Writing: The Case of Fan Fiction. In Mark Sebba, Shahrzad Mahootian & Carla Jonsson (Eds.). Language Mixing and Code-switching in Writing: Approaches to Mixed-Language Written Discourse. London: Routledge, 233-254.

Moll, Andrea. 2014. Authenticity in dialect performance? A case study of “Cyber-Jamaican”. In: Indexing authenticity: Sociolinguistic perspectives. Véronique Laoste, Jakob Leimgruber & Thiemo Breyer (eds.) Berlin/New York: de Gruyter: 209-243.

Nguyen, Dong, Dolf Trieschnigg and Leonie Cornips. 2015. Audience and the use of minority languages on Twitter. Proceedings of the Ninth AAAI International Conference on Web and Social Media (ICWSM-2015) p. 666-669.

Nguyen, Dong and Leonie Cornips. 2016. Automatic Detection of Intra-Word Code-Switching. In: Proceedings of the 14th SIGMORPHON Workshop on Computational Research in Phonetics, Phonology, and Morphology. Berlin, Germany: Association for Computational Linguistics: 82—86.