piwik-script

Intern
Geolingual Studies

Current projects

We are developing a transdisciplinary mixed-method framework to study the dynamics of urban discourses and transformations via a set of spatial and linguistic analyses. Our projects vary in their locations from global megacities to fast-growing medium-sized cities, and rural and remote areas. 

 

Within large urban spaces, sub-communities emerge and dissolve as residents move to, from and within the urban space. These developments have an impact on individual and communal identities, which can be expected to become visible in language use on structural and discursive levels (Meyerhoff & Walker, 2007). When talking about places in the city, speakers reveal and negotiate their individual thoughts, attitudes, and emotions towards them, thereby constructing place identities which are influenced by personal memories, stories and experiences. To address such (potentially changing) urban place identities in Helsinki, Huovinen et al. (2017) introduces methodological approaches for urban place studies.

This project aims to contribute to this emerging field of research focusing on two (mega)cities, namely Davao City in the Philippines and Hong Kong. It takes a comparative approach to explore how urban place identities are constructed across culturally diverse communities in distinct locations. Using interviews and a web-based questionnaire, residents are asked to describe and evaluate their city overall as well as selected districts which differ geographically and demographically. The ratings of districts are quantified and the responses to open questions analysed with a discourse analytical approach to identify relevant topics and make visible the thoughts and emotions of the newly arrived as well as the views of the established elites, their sense of belonging and their construction of hybrid identities. To analyze the participants’ narratives, the Appraisal Theory (White & Martin, 2015) is chosen as a framework.

People involved in the project

2 English linguistics, University of Würzburg, Am Hubland, 97074 Würzburg, Germany.

During the Covid-19 lockdowns, social media platforms became essential tools for socially distanced communication. Drawing on research conducted in four neighbourhoods in London, UK, the paper investigates how ‘places’ are inherent features of diverse neighbourhoods during the Covid-19 (Cowie et al., 2022).

In this paper, we ask how places shape users’ identities during three phases of Covid (start-peak-current), across diverse London neighbourhoods. Using topic modelling and a bottom-up approach (Charmaz, 2006; De Fina, 2022), we examine how Twitter users discursively engage with place and time, which index their ideologies towards migrants and their socio-economic locations.

Preliminary results indicate that users from high economic areas showed relatively increased engagement with topics surrounding place and migration during lockdown. For place, tweets featuring keywords associated with special events (e.g., concert, match) and local places of interest (pub, school) drove the increase. For migration, engagement sharply increased after January 2021, when the UK withdrew from the EU.

Further analysis combines survey-based and data-rich approaches to understand how users identify with communities and places during and after the Covid-19 lockdowns. We provide the basis for the construction of a novel multi-level place index that significantly contributes to the study of interface between place – media – ideologies before, during, and after the pandemic.

People involved in the project

2 Department of Linguistics and English Language, Edinburgh University, Dugald Stewart Building, 3 Charles Street, Edinburgh, EH8 9AD, United Kingdom

Presentations and publications

News media outlets continue to inform about the ongoing humanitarian crisis in Western Africa. Thousands of refugee encampments in this region involve numerous countries and millions of people and are a constant security concern. Many of these remote and difficult-to-access encampments remain challenging for reporting refugee-related developments and incidents in conflict-ridden areas, often leading to limited content or misplaced media attention.

This project systematically studies (1) the co-occurrence of linguistic patterns in news coverage and (2) distinguishes between the types of news offered. A dataset that comprises 2,017 refugee-related news articles annotated for their spatial and thematic dimensions will be used in this study. We will pursue a multi-dimensional analysis that identifies clusters of linguistic features that can be summarised into dimensions of register variation.

By analysing co-occurrence patterns in the news and correlating these with the situational context of their production and publication, we will identify functional relationships between the two. With this approach, imbalances in news distribution across locations can be revealed, allowing us to better understand the relations between the ‘newsworthiness’ of the regions and reporting biases linked to them.

People involved in the project

1Earth Observation Center, German Aerospace Center, Oberpfaffenhofen 82234, Germany.

Presentations and publications

The use of big data to analyze urban discourses is a growing trend. In this context, Twitter represents a valuable data source, due to its global coverage and free access. Compared to traditional ways of retrieving social data via surveys and censuses, Twitter provides past and real-time information in big data volumes with a low collecting effort. In this work, we harvested all tweets from New York City during 30 December 2021 and 2 January 2022 to analyze the intra-urban diversity of topics and their spatial distribution. We applied multilingual topic modelling to classify Tweets in different languages and spatial analysis to assess their relevance in different areas of the city. Results of this work are used to discuss the potential of such an approach, with the aim to encourage scholars from linguistics, geography and other disciplines to collaborate to assess urban discourses in a multidisciplinary manner.

People involved in the project

1Earth Observation Center, German Aerospace Center, Oberpfaffenhofen 82234, Germany.

Presentation and publications

This project examines the city discourse of Edinburgh, Scotland, in six different wards during three timepoints of the Covid-19 pandemic. Without involving in the discussion about the pandemic, we reevaluate the city’s places during the lockdowns. We combine spatial and linguistic analyses to offer a more thorough approach to 8,490 geolocated tweets via topic modelling, qualitative thematic analysis and appraisal framework, which explores the language of evaluation, attitude and emotion.

Mapping the discourse to physical places, we tackle the city’s complexity visible through the heterogenous perception and construction of places. By focusing on the people’s daily communication on Twitter, we identify the role of places in the city in relation to the spatio-temporal characteristics of the wards.

The preliminary findings show the varying topic distribution across wards (e.g., green space, vaccination, politics). Evaluation of discourse about places revealed strong attachments and positive appreciation of local places.

People involveed in the project

Presentations and publications

Monitoring and understanding current human migration patterns is challenging, particularly in crisis situations. Twitter data represents a highly valuable source to cope with this, providing real-time data including users’ information, geolocation, and opinions. Nevertheless, the Twitter discourse of migrants at different migration stages has not been explored. In this work, Twitter data was used to analyze the mobility and topics discussed by the migrants who left Ukraine after the start of the war on 24 February of 2022. Bots, mobility and text filters were applied to control the reliability of the detected displacements and the text content. Furthermore, a multilingual neural topic modelling framework was implemented to identify the discourse that migrants raised on Twitter. Based on Twitter users’ migration stage, the topics discussed by migrants before leaving, after leaving and after returning to Ukraine were assessed. The results of this research show that the integration of spatial analysis and natural language processing techniques applied to social media data is a promising approach to better understand crisis such as migration triggered by social conflicts. This information can potentially complement humanitarian response plans, adding the voices of migrants at diverse migration stages to the official information and surveys in near real-time.

People involved in the project

1Earth Observation Center, German Aerospace Center, Oberpfaffenhofen 82234, Germany.

Presentations and publications