The Secondary Archive: Collective Memory, Mental Images and AI-Assisted Representations in a Transforming Tirana

Abstract
Tirana, the capital of Albania, has experienced an intense phase of urban transformation over the last two decades. Large-scale redevelopment, densification and the redesign of public spaces have significantly altered the physical configuration of the city and, consequently, the ways in which its inhabitants perceive and remember urban places. This paper investigates how collective memory can function as a secondary archive for understanding rapidly transformed urban environments. Focusing on selected central areas of Tirana that have undergone significant physical reconfiguration, the research examines the relationship between contemporary architectural interventions and residents’ memories of the same urban fragments. Grounded in theories of mental mapping, mental representation and collective memory, the study considers individual recollections not as subjective accounts external to architectural knowledge, but as spatial evidence capable of revealing forms of continuity, disappearance and transformation that are not necessarily contained within official archives. Memory-based narratives, drawings and mental maps are therefore proposed as primary research material and are compared with historical archival sources and the contemporary urban condition. The paper further proposes an AI-assisted methodology for processing and relating these heterogeneous sources. Rather than replacing historical documentation, artificial intelligence is understood as an analytical and generative instrument capable of identifying recurring spatial references and visualizing relationships between documented and remembered environments. The resulting framework contributes to a broader understanding of urban resilience by positioning memory, identity and belonging as dimensions of the built environment that should be considered alongside physical transformation.
Keywords: Memory, Image, AI, Heritage