Student speakers

Saturday, April 23rd 2022

Sofiya Ros (11:50 - 12:10)

Sabar drumming: music or language?

In some African cultures, drumming is used for expressing linguistic meanings. Our research focuses on the Sabar drumming tradition from Senegal, which involves the encoding of spoken phrases into drum music. Playing the sabar involves at least 9 different drum strokes (hand strokes, stick strokes or their combination), that can be seen as the basic units of the Sabar language. These strokes compose different, longer strings of strokes which correspond to the spoken utterances in Wolof — the lingua- franca of Senegal. The main goal of the research is to find mapping between the drum language Sabar and the spoken language Wolof.

During fieldwork in Senegal an extensive Sabar dataset was collected. The analysis has shown that different drum strokes are more commonly associated with different types of vowels, where the association relies on the phonological properties of a vowel. In summer 2021 we conducted field research with the Senegalese drummers in Barcelona and in Lille (France). Some differences were noticed between the praxis of previously recorded drummers in Senegal, and the ones living in the diaspora. Drummers in Europe were playing only the rhythmic phrases, both traditional and modern, however, all learnt by heart beforehand, so they didn’t improvise. When they were asked to drum a specific phrase, different drummers used different strokes sequences to represent it, while the only consistency was the amount of strokes, coinciding with the number of syllables in a phrase. What concerns the spheres where the sabar drumming is used, in Europe it is now mostly used in workshops and concerts, while in Senegal it appears in different sorts of events such as sport events, life-cycle ceremonies, political gatherings.

Most likely here we deal with the fossilisation of Sabar as a language or rather it’s turning from language to music.

Bertram Mullin (11:50 - 12:10, presenting online)

English as a Foreign Language Education Beyond University: Pronunciation Studies

Recent second language (L2) pronunciation articles have neglected to address all L2 speakers’ needs in pronunciation corresponding with proficient phonetic sound structure, especially in respect to where adult learners are concerned. In this current study, 21 first language (L1) Japanese speakers learning English as an L2 of three English proficiency levels read sentences related to segmental and suprasegmental targets in order to train in phonetic speech through organized pronunciation training. A four-month examination process using Praat graphs, repeated spoken assessments, and interviews were conducted to investigate the methods learners comprehended spoken discourse of individualized problematic phonemes through three post-assessments. Data from the assessments, graphs, and interviews revealed that influences causing spoken utterance problems were due to the L1 phonological impact, L2 transfer alterations, and preconceived misinformation. The project results revealed that if participants were inactive and untreated the difficulties that they had continued as indicated in three post-assessments overtime. Apropos, the results alternately challenged that, participants who were active in the study and received two stages of treatments containing high-variability pronunciation training (HVPT) with organized sequences of pronunciation tasks overtime, including pronunciation comprehension tasks that were listening-test based, in L2 spoken discourse could be assisted with difficulties in phoneme comprehension.

Elise Alberts (14:40 - 15:00)

Religion in a Second Language: Language Attitudes in International Anglophone Churches in the Netherlands

Something remarkable is happening in secularised the Netherlands: although Dutch-speaking churches are in decline, international English-speaking churches are drawing in many speakers with Dutch as their first language. These Dutch speakers made a choice to join a religious community in their second language instead of their first, while the Christian faith is often claimed to be personal and emotional and thus better experienced in a first language (Brown, 2009; Burke, 2010; Dye, 2009). The focus of this study is on the attitudes of Dutch people towards the use of English in their religious life. Through 13 interviews with church leaders of international anglophone churches in the Netherlands and a survey among 95 Dutch people who frequent them, the study showed that the Dutch language was viewed as an unattractive language in the religious sphere, while English held a certain kind of aesthetic quality that was seen as more suitable to express religious thoughts and ideas. This was especially the case when it came to religious music, as a large majority preferred to sing in English instead of Dutch. In addition, several participants experienced the English religious language as more natural and comprehensible than the Dutch religious language. This study showed that the use of English is increasing in Dutch society, beyond the large variety of domains already mentioned in earlier research by Edwards (2015), and that it is permeating in the religious domain as well.

Paula Badilla & Gabriel Colipán (14:40 - 15:00, presenting online)

Testing Linguistic relativity: perceiving the world through more than one language

Over the last few decades, there has been a renewed interest among language scholars around the hypothesis of linguistic relativity. Since this idea deals with difference in perception between speakers of different languages, bilinguals have been common participants on linguistic relativity studies. However, their mother tongue is usually the target of these studies; their other languages remain only as a common language among other participants. Bilingualism per se has, to our knowledge, not been the main focus of these studies. This research thus aims to test the precepts of Linguistic Relativity among Spanish (L1)-English (L2) bilinguals in Chile, inquiring whether the activation of only one or both of their languages—also known as language modes—correlates with a change in bilinguals’ perception of narrated events. To this end, twelve Spanish dominant Spanish-English bilinguals were selected. Six groups of three participants each were required to interpret a text and three images. Four were groups of bilinguals in which either a monolingual or a bilingual mode of language was triggered by making them process information and communicate either in one single language or in both simultaneously. The additional two were control groups of Spanish monolinguals, and English native speakers. The text consisted of a diary entry of a person, whose gender, age, and general personality were left to the participant’s interpretation. The images showed characters in different situations, which participants were asked to describe and to create a narration with. Participants of the six groups individually carried out the same task, based on a heavily linguistic stimulus as well as on an iconic (non-linguistic) one. Results thus reveal that not only the language bilinguals choose to speak, but also the mode of language in which they are in, seem to indeed influence the aspects of reality that are perceived and highlighted while discussing events narrated in a certain language.

Jeroen van Ravenhorst (15:15 - 15:35)

Comparison of semantic structures between Surinamese creole and West African languages

Sranan Tongo is an English-based creole that emerged as a contact language between the enslaved African population and the European plantation owners in northern Suriname during the second half of the 17th century. Of these enslaved African people, most had their origins in the Slave Coast. This area corresponds to modern day Togo and Benin, a predominantly Gbe-speaking area. For this reason, Sranan is assumed to have large Gbe substrate influence. This is mostly based on morphosyntactic, lexical, and phonological evidence.

In this presentation, I take a semantic approach to the investigation of African substrate in Sranan. I compare the Sranan's semantic structures to four West African languages: Ga, Twi, Ewe, and Kabiye. This study builds heavily on an earlier study by Huttar et al. (2007), which compared 101 senses of Ndyuka (Eastern Suriname Creole) nouns, verbs, and other lexemes with senses of corresponding lexemes in those four languages. Comparing the results from semantic structures attested in Ndyuka with new data from Sranan sheds light on the historical and linguistic connection between Suriname and particular West African coastal areas characterized by the presence of these respective languages.

For example, compare the correspondence of the following semantic structures between the languages:

HUNGER KILLS N = ‘N is hungry’

angri kiri N

> Sranan, Ewe

MEAT/FLESH = ‘animal’


> Sranan, Ga, Ewe

BREAST MOUTH = 'nipple'


> Sranan, Ga, Twi, Ewe, Kabiye

Sunday, April 24th 2022

Matthijs Verzijden (11:50 - 12:10)

Hakka mountain songs

In this talk, I investigate san24go24 ‘mountain songs’, an oral tradition of the Hakka. This linguistic, cultural, and maybe ethnic group, living in the south-east of China, in Taiwan, and in the diaspora, stands out from other Chinese ethnolinguistic groups for its strong group identity and allegedly unique history. Although Hakka identity is indeed strong and exclusive, it is also constructed and more recent than is commonly argued (Constable 1996; Leong 1997). I argue that the san24go24 oral tradition can not only illustrate these issues very well, but may also bring some new insights to this discussion.

Through the analysis of one performance of these mountain songs from Taiwan, compared with the findings from scholarly literature, I investigate the general properties of this oral tradition as for content, form, music, and performance. San24go24 tell of love, desire, and sex; they feature rigid verse form, meter, and rhyme; melodies vary only limitedly between the songs; the songs are primarily sung by women and may be accompanied by a small orchestra, but much has changed throughout the tradition’s history (compare Idema 2015). These observations enable me to connect the tradition with its sociocultural context, with other similar oral traditions around as well as Sinophone literature at large, and with issues in oral tradition description and analysis in general.

It turns out that, while Hakka mountain songs are argued to be uniquely Hakka, and do indeed exhibit many Hakka cultural features, they also show some remarkable similarities to other Sinophone but non-Hakka oral traditions. In addition, the history of Hakka mountain songs shows how oral literature is never entirely separate from written literature. Finally, I show which role filler syllables and monothematism, as issues of oral tradition study at large, play in Hakka san24go24, and how these two topics provide avenues for further study.

Wout Damen (11:50 - 12:10, presenting online)

Handshape acquisition in Sign Language of the Netherlands

Signs are made up of five phonological parameters, Handshape being one of them. Children acquire handshapes in a semi-predictable order, comparable to phoneme acquisition, starting with common unmarked handshapes and ending with the marked ones. The exact order of handshape acquisition has been examined for some sign languages, but not yet for Sign Language of the Netherlands (NGT). My BA thesis is exploratory study presenting an initial acquisition order for NGT handshapes based on longitudinal data from a child with NGT as his mother tongue. The orders derived from this data have been analysed with a more traditional method in the study of handshape acquisition and an alternative method from phoneme acquisition. Analysis of the orders derived from these two methods argue that the alternative method has a higher validity and would therefore be suitable for further analyses of sign language acquisition. The final acquisition order of NGT handshapes can be accounted for in terms of markedness and feature acquisition. Potential universals of the sign language acquisition process based on this study are /1/ and /5/ as fundamental handshapes, the early acquisition of the unmarked handshapes, and the acquisition burst of marked handshapes.

David Ginebra (12:25 - 12:45)

Segments defining the South America macro-area: evidence from a quantitative study

South America (hence SA) is currently considered a linguistic macro-area, that is, the sum of different language regions (Campbell 2012). Efforts have been made to find common traits among the languages in these areas (Wichmann et al. 2011; Campbell 2012; Aikhenvald 2012). This paper aims to contribute to the description of this area by analyzing shared phonological traits: I point out segments which are common in SA but rare elsewhere (hence EW), and segments which are common EW but rare in SA.

This study takes a quantitative approach, consisting of a systematic analysis and comparison between the SAPhon and PHOIBLE databases is made for the first time. SAPhon is a database including inventories for 363 SA languages (Michael et al. 2015), while PHOIBLE is a database including inventories for 3020 world’s languages (Moran et al. 2014). Combining data from both databases I have calculated how often each segment is found in SA and EW and the function between these two values. For instance, the velar nasal segment /ŋ/ is found in 22% of SA languages and in 68% of the languages EW. The function between both values is 3.14, that is, /ŋ/ is +3 times more frequent EW than in SA.

The results of this study are threefold: first, they show new phonological characteristics of SA languages; second, they reinforce some assumptions in the description of SA languages; and finally, they clarify some previous misconceptions. Nasal vowels and the high central vowel /ĩ/ appear to be significantly more common in SA than EW. This is also the case for the alveolar flap /ɾ/ (80% of the languages in SA but only 18% EW); in contrast, the alveolar trill is extremely rare (only 4% of SA languages). Labiodental fricatives are also very rare: /v/ is only found in 1% of SA languages and /f/ in 5% (18 and 9 times less frequent than EW, respectively).

These findings show some common characteristics in SA language inventories, but they also show a quantitative analysis approach that might be used for further research, such as assessing language areas according to their phonological similarities in SA and also EW.

Jelle Christiaans (12:25 - 12:45, presenting online)

Reinventing a 2500-year-old category: on verbal aspect, breaking with tradition, and persuasive rhetoric (no knowledge of Classical Greek or the Greek alphabet necessary!)

Even after two-and-half millennium, the jury is still out on the precise workings of verbal aspect in Classical Greek and the semantic value of the aorist (perfective aspect) and present (imperfective aspect) stems in particular. The general consensus is that perfective aspect presents completed events and imperfective aspect not yet completed events, but classicists tend to gloss over cases where these values do not hold up. A total review of the category of aspect is needed, but this is no easy feat. In this talk, I will take you along the journey of how I, together with Dr. Adriaan Rademaker, approached this revamp of Greek verbal aspect. Obviously, it will be about Greek, but my bigger aim will be to sketch the way we work in order to give you some insight into how academic puzzles may be laid.

Be prepared to meet an obstinate professor of Greek no-one ever believed, to encounter structuralist linguists stuck in the previous century, and to relive an ancient court case and a divorce quarrel. During the talk, I will guide you through the communis opinio, the problems it presents and the different bits of data we collected to piece together the puzzle of Ancient Greek verbal aspect. We will pass by old ideas, rebellious souls and clever (and not-so-clever) refutations to end up at the final conclusion: Classical Greek verbal aspect is best understood as a complex, polysemous category that can be described using a network of uses.

Ariëlle Reitsema (12:50 - 13:10)

Chameleon or Copycat? The Morphological Causative and Passive in Korean

Foreigners studying Korean grammar are bound to come across a peculiarity: causative and passive verbs can often have identical surface forms. For example, the verb poita – derived from pota “to see” – can mean either “to show” or “to be seen” depending on the context. At least in the surface representations, it appears that exactly the same suffix –i is used to create both the morphological causative and passive. Although this does not tend to lead to misunderstanding in daily life, the ambiguity is striking and remains to be accounted for.

My paper focused on how this I-suffix should be represented in the grammar: as one or two. Are there in fact two separate underlying morphemes that are like copycats and have therefore grown more alike, or is there one underlying suffix that is versatile like a chameleon and can somehow ‘change its colour’ to fit in multiple different contexts?

There were proponents of both approaches in the literature. Those supporting the chameleon-approach represented only one suffix, drawing arguments from the doubling constraint and adversity constructions. (A passive causative such as “food was fed to her” cannot be created by doubling the I-suffix, although it can be formed differently.) Proponents of the copycat-approach on the other hand argued that the selection of the passive and causative suffixes is lexicalised and therefore separate. (For example, some homonyms select a different phonological variant of the I-suffix for the passive and causative forms respectively.)

A critical examination of the arguments showed some weaknesses in the argumentation in the copycat-approach, and underlined the value of a diachronic perspective in explaining patterns despite irregularities. Thus, the I-suffix was analysed as having a single underlying representation, while keeping in mind that the suffix seems to be intrinsically ambiguous between being one or two suffixes altogether.

Kambiz Nasseri (15:00 - 15:20)

The effects of individual differences in L1-complexity on L2-complexity

Previous research indicated that individual variability in L1-profiency can carry over to the L2. This has implications for the notion of an ultimate native speaker level of L2-attainment, as it suggests that (1) native speakers differ with regard to their proficiency and (2) a speaker’s L2-performance might (partially) be a product of their individual speaking style rather than the level of L2- attainment. This in turn has implications for assesments of L2-speech, as well as theories of L2- learning. Together with a group of students, I investigated whether such an effect could be found for complexity of spoken speech. 24 participants with various (linguistic) backgrounds performed four speaking tasks, two in the L1 and two in the L2. Four measurements were derived from the data, two of which measured syntactic complexity (‘mean number of clauses per AS-unit’ and ‘mean number of words per clause’) and two for lexical complexity (‘mean word length’ and ‘Guiraud index’). Spearman rho tests revealed that individual variability in L1-scores was positively correlated with L2-scores for all measures. This suggests that the evidence for the effect of individual differences in L1-profiency on L2-profiency can be extended to the domain of complexity of speech. Additionally, I investigated whether the level of prior education could be identified as a predictor for both L1 and L2-scores, but this effect was not found.

Aholi So (15:35 - 15:55)

Language Transmission Challenges of Cantonese Heritage Speakers in the Netherlands

As the Dutch society becomes increasingly more aware of the different needs of people with a migration background, the question of heritage language transmission becomes more relevant as well. This is especially evident amongst heritage speakers of different language varieties. Second-generation Cantonese speakers in the Netherlands, whose dominant language is generally not their heritage language, experience difficulties in finding a balance between family language planning and their own language competence. In this presentation, I discuss the challenges Cantonese heritage speakers foresee and experience in the transmission of Cantonese to the next (third) generation Cantonese. This presentation draws on the data I collected from intensive 1-on-1 online semi-structured interviews during the COVID-19 pandemic from September to December 2021 and is part of my ResMA thesis about Cantonese as a Heritage Language in the Netherlands. I will discuss the fragility of heritage languages, while disentangling the thoughts, language attitudes and understandings and misunderstandings among Cantonese heritage speakers concerning heritage language maintenance and transmission. In addition, I discuss the practicalities of heritage language transmission, in which there is a role for parents, grandparents, Cantonese schools and educational institutions more generally.

Got a question?

Don't hesitate to send us an email at

Also check out our Instagram , Facebook and LinkedIn page!

Who organizes the Conference?

The conference is organized by the Conference Committee of Studievereniging T.W.I.S.T., the study association for Linguistics at Leiden University.