Cambridge Language Sciences Annual Symposium 2018

Language Sciences and Health

Now in its fourth year, this annual meeting of minds brings together established and early careers researchers from the University of Cambridge for an afternoon of talks and poster presentations, and provides an opportunity for informal networking.

This is an internal event free to Cambridge researchers. If you are a researcher from another institution or a Cambridge undergraduate and would like to attend, please contact the organiser.

**Please note: talks by the speakers will be filmed, but members of the audience will not**

Programme

13.00-13.30 Registration and coffee; poster exhibition

13.30-14.30 Keynote lecture: Language acquisition, neural entrainment, phonology and dyslexia

Professor Usha Goswami (Centre for Neuroscience and Education, Dept. of Psychology)

Recent insights from auditory neuroscience provide a new perspective on how the brain encodes speech. Using these recent insights, I will provide an overview of key factors underpinning individual differences in children’s development of language and phonology, providing a context for exploring atypical linguistic development. I will develop an oscillatory “temporal sampling” neural framework for linking rhythmic auditory processing to linguistic development by children. I will show that sensitivity to the amplitude modulation (AM) structure of infant-directed and child-directed speech is key to individual differences, and that this AM structure contains acoustic statistical cues to phonology and morphology. Children with dyslexia are relatively insensitive to amplitude modulations and rhythm patterns. I will show that their lack of rhythmic sensitivity is related to the atypical neural encoding of energy patterns in speech via neuronal oscillatory entrainment, and I will explain how this mechanistic process is atypical in dyslexia, offering new avenues for potential remediation via assistive listening technology.

14.30-15.00 Health crises, digital media and community voices: utilising interactive radio for rapid social research to improve outbreak preparedness and response

Dr Claudia Abreu Lopes (Affiliated Lecturer, Dept. of Politics and International Studies, and Research Advisor for the project Digital Media and Health Communications in Africa)

In 2017 Somalia experienced the worst cholera outbreak in five years, with over 79,000 cases and over 1,100 deaths mainly among children under five years. Longitudinal analysis of case reports confirms that despite many years of public health interventions cholera remains a recurring and major risk to vulnerable communities in the country.

This project aims to evaluate a tool for understanding cholera risk and preparedness among communities in the South Central Zone of Somalia. Such an understanding could both provide an early warning system, as well as point to specific socio-cultural elements that could influence the success of an intervention during a cholera outbreak.

Specifically, the project considers an interactive radio and SMS-based method as a means for data gathering. This presentation will focus on the methodology that Africa’s Voices Foundation employed to gather, label and analyse the SMS data, as well as lessons learnt and how these could support both programmatic interventions and other similar data gathering initiatives in the future.

15.00-15.30 Poster slam (1-minute talks by the poster presenters)

15.30-16.30 Poster exhibition; refreshments

16.30-17.00 Using NLP and heterogenous user generated content to sense mental well-being

Dr Maria Liakata (Dept. of Computer Science, University of Warwick)

A large body of work in natural language processing (NLP) for clinical applications is based on processing electronic health records (EHRs). While the latter are rich in information there are typically only few records per patient. More recently there has been interest in processing user generated content (UGC) such as social media posts collected over time to make predictions about individuals' mental health. Such UGC data is available at much more frequent temporal intervals than EHRs but may be noisier. So far the majority of work in NLP on mental health prediction, even when using longitudinal social media data, involves distinguishing individuals with a condition from controls rather than assessing individuals’ mental health at different points in time. Some work in the area of mobile computing has used sensor information to make longitudinal predictions but this hasn't considered any language content. I will present our work on integrating both linguistic data and heterogeneous and asynchronous mobile phone usage data for monitoring mental health over time. I will also discuss the challenges such work entails when evaluated in a challenging scenario for real world deployment.

17.00-17.30 The effect of early language and communication environment on social outcomes for primary school aged children with language difficulties

Dr Jenny Gibson (Faculty of Education)

Children who have difficulties in linguistic development that persist into middle childhood are at higher risk of later social problems, in comparison to their typically developing peers.

In this talk I will present findings from a study based on data from the Avon Longitudinal Study of Parents and Children (ALSPAC). The study aimed to explore the extent to which a child’s early communicative environment may influence later social adjustment, and, whether the pathways for such influences were different for children with and children without language difficulties. I report an analysis using a subset of ALSPAC participants who completed direct assessments of linguistic/psycholinguistic abilities in middle childhood.

394 children who had evidence of language difficulties (LD) at age 7 years were identified, alongside a sample of typically developing (TD) children matched for age, gender and SES (788 children in total, 59% male, 50% with language difficulties). Pathways between early communicative environment (18-24 months), play and social development in middle childhood (7-9 years), and, later social difficulties (11 years) were compared between the groups.

Whilst for both groups of children, being prosocial and engaging in play were protective against later behavioural problems, for children with LD, these protective effects were more pronounced for those who had a more positive early language and communication environment. I will discuss these findings with reference to support and education strategies for children at risk of language difficulties.

17.30-18.30 Keynote lecture: Stability and change in child language

Professor Courtenay Norbury (Literacy, Language and Communication Lab, University College London)

Longitudinal studies of child language development and disorder demonstrate both steady language growth yet remarkable stability in the rank order of language competencies, at least from about the age of 4. The extent to which growth and stability are influenced by other aspects of child development is a focus of the current study.

In this talk I will present data from the Surrey Communication and Language in Education Study (SCALES), a population study of language change and stability from school entry. We followed approximately 500 children with a diverse range of language, cognitive, and social/emotional abilities from Reception to Year 3 (ages 4-8 years).

All children showed growth in language skill, yet language was incredibly stable (ICC = .95). Thus, children with relatively low language scores at the beginning of formal education continued to have relatively low language scores in Year 3. Importantly though, children with multiple developmental challenges were not falling further behind, at least in mid-primary school. In fact, the rate of language growth was remarkably similar in three groups of children with diverse language and cognitive profiles. These findings raise important questions about the nature and ultimate goals of intervention programmes for children with language disorder.

Award of the poster prize; closing remarks

18.40-19.45 Reception

Drinks reception supported by Cambridge University Press

Event image photo credit: Kaili Clackson (copyright image not for reproduction)

Posters

1. A collaborative game-based approach to documenting linguistic variation in Brazil
Jenna Renjie Zhou, Andrew Caines, Paula Buttery, and Ioanna Sitaridou
*Research supported by the Language Sciences Research Incubator Fund and the Isaac Newton Trust

2. Developing a prototype web-app for numeracy assessment and teaching Zhilin Wang, Andrew Caines, Russell Moore, Paula Buttery, and Ianthi Tsimpli
*Research supported by the Language Sciences Research Incubator Fund and the Isaac Newton Trust

3. Developing prototype assistive listening technology for remediating developmental dyslexia Sheila Flanagan, Anji Wilson, Fiona Gabrielczyk, and Usha Goswami

4. Functional distributional semantics: learning linguistically-informed representations Guy Emerson and Ann Copestake

5. Historical codeswitching and language mutability in the history of English Marieke Meelen and Laura Wright
*Research supported by the Language Sciences Research Incubator Fund and the Isaac Newton Trust

6. How clever are the models exhibiting ‘super-human’ performance on VQA datasets?
Huiyuan Xie, Alexander Kuhnle, and Ann Copestake

7. Indefinite NPs as subjects in L2 and L3 Mandarin grammars Jingting Xiang and Boping Yuan

8. Indirect corpus applications: collecting data and using it in ELT materials development Niall Curry and Olivia Goodman

9. Inter-symptom relationships across linguistic, cognitive, and behavioural problems in a developmentally at-risk population: a network approach Silvana Mareva, Joni Holmes, and the CALM team

10. Investigating mechanisms of change in children’s social-attributional thinking: a frame semantics approach to sociocultural discourse analysis Courtney Froehlig

11. Language, Brains & Machines: an initial literature review Anna Samuel, Andrew Caines, and Paula Buttery
*Supported by the Language Sciences Research Incubator Fund and the Isaac Newton Trust

12. Language processing of personal beliefs Kaili Clackson, Tristan Bekinschtein, and Valdas Noreika

13. L2 English genitive choices of L1 Spanish speakers James Algie

14. Matrix language variation in a corpus of Vietnamese-English code-switching in Canberra Li Nguyen

15. Mobile-based Dictionaries (MBDs): investigating Chinese EFL learners’ use and the effectiveness in incidental English vocabulary learning Danyang Zhang

16. Modelling lexical decision in adults with and without acquired dyslexia and semantic dementia Ya-Ning Chang, Matthew Lambon Ralph, and Stephen Welbourne

17. Modelling polysemy in distributional semantic models Guy Aglionby, Helen Yannakoudakis, Jean Maillard, and Ekaterina Shutov

18. Modulating speech processing with alternating current stimulation Benedikt Zoefel, Isobella Allard, Megha Anil, Alan Archer-Boyd, and Matthew H. Davis

19. Multilingualism and wellbeing: a scoping review Lisa-Maria Müller, Katie Howard, Elspeth Wilson, Jenny Gibson, and Napoleon Katsos
*Supported by the Language Sciences Research Incubator Fund and the Isaac Newton Trust

20. Multilingualism, multiculturalism and implications for health and wellbeing Özge Ozturk, Julija Vaitonyte, and Lisa-Maria Müller

21. Multimodal semantic revision during inferential processing: the role of inhibitory control with story and picture information Ana Pérez

22. Multiple sentence representations in machine translation Danielle Saunders, Felix Stahlberg, Adria de Gispert, and Bill Byrne

23. Neural coding schemes for lexically-driven prediction in superior temporal cortex
Ediz Sohoglu, and Matthew H. Davis

24. Predictive neural mechanisms of spoken word recognition and learning Yingcan Wang, Rebecca Gilbert, Richard Henson, and Matthew H. Davis

25. Semantic memory performance after left vs. right anterior temporal lobe resection Grace Rice, Helen Caswell, Perry Moore, Paul Hoffman, and Matthew Lambon Ralph

26. The effects of language contexts on executive function: a pilot study Orianna Bairaktari, Eva Periche-Thomas, Andriani Papageorgiou, Antonella Sorace, and Roberto Filippi

27. The neural representation of colour knowledge entails brain regions beyond the visual cortex Rocco Chiou and Matthew Lambon Ralph

28. The school experiences of bilingual children on the Autism Spectrum Katie Howard, Jenny Gibson, and Napoleon Katsos

29. Towards a computational model of controlled semantic cognition Rebecca Jackson, Timothy Rogers, and Matthew Lambon Ralph

30. Towards automatic grammatical error detection of non-native spoken learner English Kate Knill, Mark Gales, Potsawee Manakul, and Andrew Caines

31. Why are Asians better in mathematics? A modern investigation into linguistic relativity on number counting Ann Dowker and Anthony Li