Between Blad (Homeland) and Ghorba (Exile): An NLP-Assisted Semiotic Analysis of Cultural Identity Construction in Algerian Migrant Discourse

Abstract

Migration constitutes both a socio-economic phenomenon and a discursive process through which cultural identities are continuously constructed and renegotiated. In Algerian migrant communities, the Darija terms blad (homeland) and ghorba (exile) function as central semiotic operators that organize experiences of belonging, displacement, memory, and affective rupture. Despite their prominence in everyday discourse, systematic analysis of how these terms structure cultural identity remains limited. This study investigates how blad and ghorba are discursively constructed in Algerian migrant narratives, identifies their semantic and collocational patterns, and evaluates the contribution of AI-assisted corpus analysis. Adopting a convergent mixed-methods design, the research combines quantitative corpus analysis of 1,248 textual units collected from three Algerian diaspora Facebook groups in France with ten semi-structured qualitative interviews with Algerian migrants residing in France. Corpus processing was conducted using spaCy (v3.7) with custom normalization for Darija–French code-switched text, including frequency counts, collocation analysis via normalized pointwise mutual information (NPMI), and qualitative semiotic interpretation. Results show that blad is consistently associated with family, memory, warmth, and cultural authenticity, while ghorba is linked to loneliness, suffering, and spatial-cultural dislocation in France. Interview data reinforces this semiotic opposition while revealing reflexive hybrid negotiation and platform-specific identity performances. The study demonstrates both the value and the limitations of AI-assisted methods for preserving cultural depth in migration linguistics and critical discourse analysis.