The thesis concerns the linguistic landscape (LL) of two neighbourhoods, one in Pretoria, South Africa, and the other in Marseille, France. This is a longitudinal study whose data was collected over two years of site visits. LL are explored in terms of both space and place. In terms of place, they are seen to be constitutive of a sense of place, allowing insights into memory, aspiration, and familial and cultural networks. Spatially, they are seen to realise a politics where design and distribution of LL are markers of power and modality. Analysis takes its point of departure in geosemiotics. Artefacts of LL are interpreted as sites of encounter of four cycles of discourse: the interaction order, habitus, semiotics of place and visual semiotics. The focus is on understanding LL artefacts, their production and reception, as a nexus of practice. Methodologically, walking - as a creative practice, and as an actualisation of the place and space of the neighbourhood - is chosen for photographing LL, for observing interactions and for meeting participants to the research. In examining habitus, the discourses, literacy and narratives of the people who live, work and pass through the site are compared. Deep social and economic similarities are noted between the two sites. Exploration of the semiotics of place brings to light regularities in the features of formal and informal LL, the nature of participation with and subversion of these texts, but also disparities among producers and receivers in terms of literacy, access, the socio-cultural and the socio-economic. Visual semiotic analysis continues these findings and it is noted that global and local discourses of identification, aspiration and self-stylisation circulate transversally in the sites. LL are taken to realise a politics of space when multimodal analysis of composition and modality is extended to the streetscape, as LL ensemble. A key facet of the research is the interpretation of informal LL. Their inclusion challenges existing LL methodologies by flagging the necessity to ground quantitative findings ethnographically.