Neighbourhood linguistic accessibility for older people in Singapore: Spatial mismatch and linguistic gentrification

Linguistic accessibility, the availability of services and information in languages familiar to residents, remains largely absent from environmental gerontology and age-friendly city assessment frameworks, despite its growing significance for older adults in multilingual cities undergoing rapid language policy transitions. This study develops a framework of neighbourhood linguistic accessibility for older people and applies it to 10 planning areas with high older-adult concentrations in Singapore. Using 3922 points of interest and 31,275 online reviews collected from Google Maps, it constructs dual supply-side (D1) and demand-side (D2) linguistic diversity indices, refined through sentiment analysis to better reflect the linguistic experiences of everyday service use, and analysed through spatial clustering to identify distributional patterns across the urban environment. The results reveal a systematic spatial mismatch: linguistic diversity hotspots concentrate in the Central Business District, while mature estates where more than 25% of residents are aged 65 and above display no comparable linguistic accessibility. Within Singapore’s planning structure, which designates town centres as mixed-use daily service hubs for older residents, linguistic accessibility varies markedly, with only a minority showing meaningful multilingual diversity. These mismatches suggest two mechanisms of linguistic gentrification: globalisation-driven concentration of multilingual environments, and community-renewal-driven displacement of historically inclusive spaces. Identifying four spatial configurations and four differentiated policy categories, the study challenges the assumption that signage diversity equates to accessibility and calls for embedding language–competence alignment within age-friendly assessment frameworks, offering a transferable approach to achieving linguistic equity in multilingual urban governance.